ABSTRACT 
 
 
Title of Thesis: GEOMECHANICAL AND HYDRAULIC BEHAVIOR        
OF MARYLAND GRADED AGGREGATE BASE  
MATERIALS. 
 
 Intikhab Haider, MSc in Civil Engineering, 2013. 
 
Thesis directed by:  Associate Professor Ahmet. H. Aydilek 
              Department of Civil and Environmental Engineering. 
 
 
Maryland State Highway Administration (SHA) is in need for evaluation of 
stiffness and drainage characteristics of graded aggregate base (GAB) stone delivered at 
highly variable gradations to the construction sites.  To fulfill the current need, the 
mechanical and drainage properties of several Maryland GAB materials were evaluated 
in the laboratory and field.  The resilient modulus (RM) and hydraulic conductivity (HC) 
test results obtained in the laboratory were compared to the field RM and HC. The effect 
of moisture content on RM was also evaluated. Summary RM values at Optimum 
Moisture Content (OMC) minus 2% were higher than those at OMC, with few 
exceptions; however, the permanent deformations were increased with addition of 
moisture content. An addition of 4-6% fines over the SHA specification limit of 8% 
resulted in 2-5 times decrease in the laboratory-based GAB?s HC and an increase in time 
for 50% completion of the drainage.  
 
 
 
GEOMECHANICAL AND HYDRAULIC BEHAVIOR OF MARYLAND 
GRADED AGGREGATE BASE MATERIALS 
 
 
by 
 
Intikhab Haider 
 
 
 
 
Thesis submitted to the Faculty of the Graduate School of the 
 University of Maryland, College Park in partial fulfillment 
 of the requirements for the degree of 
Master of Science  
2013 
 
 
 
 
 
 
 
 
 
 
 
Advisory Committee: 
 
Associate Professor Ahmet H. Aydilek 
  Professor Charles W. Schwartz 
  Professor M. Sherif Aggour 
 
 
 
 
ii 
 
 
 
ACKNOWLEDGEMENT 
 
First and foremost, I would like to thank my advisor, Associate Professor Ahmet 
H. Aydilek, for his enthusiasm and guidance throughout my graduate studies at the 
University of Maryland ? College Park. Thanks also to Professors Charles W. Schwartz 
and M. Sherif Aggour for being part of my committee. Thanks to Maryland State 
Highway Administration to help us to run resilient modulus and permeability tests and 
provide us necessary information about highway construction practice that are used in 
State of Maryland. Thanks to my wife for her support during my study. 
 
 
 
 
 
 
 
 
 
  
iii 
 
TABLE OF CONTENTS 
  
ACKNOWLEDGEMENT .......................................................................................................ii 
LIST OF TABLES .................................................................................................................. iv 
LIST OF FIGURES ................................................................................................................. v 
1. INTRODUCTION ............................................................................................................. 1 
2. MATERIALS ..................................................................................................................... 4 
3.  METHODS .......................................................................................................................... 6 
3.1 Laboratory Geomechanical Tests ............................................................................ 6 
3.2 Laboratory Hydraulic Conductivity Tests.................................................................. 9 
3.3 Field Tests. ........................................................................................................ 10 
3.3.1 Light Weight Deflectometer ........................................................................... 10 
3.3.2 Geogauge.................................................................................................... 12 
3.3.3 Field Hydraulic Conductivity Tests. ................................................................ 12 
4.  RESULTS AND ANALYSIS .......................................................................................... 14 
4.1 CBR Tests. ........................................................................................................ 14 
4.2. Resilient Modulus Tests ...................................................................................... 15 
4.3. Permanent Deformation Tests .............................................................................. 18 
4.4. Laboratory Hydraulic Conductivity Tests............................................................... 20 
4.5 Field Tests ........................................................................................................ 22 
5.  PRACTICAL IMPLICATIONS ...................................................................................... 24 
5.1 Highway Base Design ......................................................................................... 24 
5.2. Effect of Hydraulic Conductivity on Highway Base Design ...................................... 25 
5.3. Cost Calculations............................................................................................... 30 
6. CONCLUSIONS ................................................................................................................ 32 
TABLES ................................................................................................................................. 35 
FIGURES ................................................................................................................................ 43 
APPENDIX A ........................................................................................................................ 73 
APPENDIX B ......................................................................................................................... 77 
APPENDIX C ......................................................................................................................... 84 
REFERENCES ....................................................................................................................... 87 
 
iv 
 
LIST OF TABLES 
 
Table 1.    Physical and chemical properties of GAB and RCA materials 
Table 2.    CBR, SMR, power fitting parameters and plastic strain of GAB materials 
Table 3a.  Effect of curing time and freeze-thaw cycles on CBR and SMR of the two    
RCAs. 
Table 3b.   CBR, SMR and power model fitting parameters of RCA/GAB mixtures. 
Table 4.     Mean hydraulic conductivity of GAB materials tested in the laboratory and 
in-situ. 
Table 5.   Effect of gradation parameters on SMR and hydraulic conductivity of GAB 
materials. 
Table 6.  Effect of change in layer coefficient and drainage modification factor on the 
required base thickness in pavement design.  
Table 7. . Effect of fines content and GAB type on required base thickness and time to 
drain. 
Table A.1.  AASHTO T 307-99 resilient modulus testing sequence for base and subbase 
materials. 
Table C.1    Field Tests Results   
v 
 
LIST OF FIGURES 
 
Figure 1- Gradation of (a) GAB, and (b) RCA materials. 
Figure 2- Gradation of field retrieved samples. 
Figure 3- GAB resilient moduli at different loading sequences 
Figure 4- Effect of moisture content on SMR values of GABs 
Figure 5- Effect of (a) fines content (b) gravel to sand ratio on SMR of Rockville GAB.  
Figure 6- Arrangement of particles in a soil matrix with the variation of fines. 
            (a)No or small fines content (large G/S ratio),  (b) dense graded (optimum 
G/S), and (c) high fines content (small G/S) (Yoder and Witczak 1975, and 
Xiao et al. 2012) 
Figure 7- Resilient moduli of recycled aggregate A and B, and their mixtures with two 
GABs. 
Figure 8- Plastic strain of GABs under repeated load cycles.  All specimens are 
compacted at OMC. 
Figure 9- Effect of moisture content on plastic strain of the GABs   
Figure 10- Plastic strain of recycled concrete aggregates A and B, and their mixtures 
with Rockville GAB 
Figure 11- Change in gradations of (a) Rockville GAB due to adjustment between 
0.6mm and 4.75 mm sieves, and (b) Rockville and (c) Texas GABs due to 
adjustment between 0.375 mm and 0.75 mm sieves.  
 
vi 
 
Figure 12- Effect of fines on hydraulic conductivity of (a) Rockville GAB due to 
adjustment between 0.6 mm and 4.75 mm sieves, and (b) Rockville and (c) 
Texas GABs due to adjustment between 0.375 mm and 0.75 mm sieves.. 
Figure 13- Effect of gravel content on hydraulic conductivity of (a) Rockville, and (b) 
Texas GABs 
Figure 14- Effect of gravel/sand ratio on hydraulic conductivity of (a) Rockville, and 
(b) Texas GABs. 
Figure 15- Effect of characteristic grain sizes on hydraulic conductivity of (a) 
Rockville, and (b) Texas GABs. 
Figure 16- Comparison of laboratory and field moduli. 
Figure 17- Comparison of laboratory and field hydraulic conductivities of GAB 
materials. 
Figure 18- Variation of required base thickness with the change in layer coefficient 
and drainage modification factor. 
Figure 19- Variation in required base thickness with base course hydraulic 
conductivity  
Figure 20- Variation in time to drain with percent drainage and fines content for 
Rockville GAB: (a) Barber and Sawyer Method, and (b) Casagrande and 
Shannon Method. H= 0.3 m was used throughout the analysis 
Figure 21- : Variation in time to drain with base thickness and fines content for 
Rockville  GAB: (a) Barber and Sawyer Method, and (b) Casagrande and 
Shannon Method.  
vii 
 
Figure 22- Variation in time to drain with percent drainage and fines content for Texas 
GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon 
Method. H= 0.3 m was used throughout the analysis 
Figure 23- Variation in time to drain with base thickness and fines content for Texas 
GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon 
Method.  
Figure 24- Influence of hydraulic conductivity on required base thickness (Moulton 
method) and time-to-drain of highway base layers constructed with (a) 
Rockville, and (b) Texas GAB. 
Figure 25- Variation in time to drain with percent drainage and GAB type: (a) Barber 
and Sawyer Method, and (b) Casagrande and Shannon Method. H= 0.3 m was 
used throughout the analysis 
Figure 26- Variation in time to drain with base thickness and  GAB type: (a) Barber 
and Sawyer Method, and (b) Casagrande and Shannon Method.  
Figure 27- Effect of GAB hydraulic conductivity on (a) required base thickness, and 
(b) time to drain at 50% drainage.  
Figure 28- Effect of time to drain and quality of drainage on the cost of GAB layer. 
Figure A.1- Particle degradation of GAB materials due to vibratory and impact 
compaction: (a) Texas (b) Rockville (c) Bladensburg (d) Churchville 
Figure B.1- (a) Resilient Modulus Testing Equipment (b) Vibratory compactor. 
Figure B-2- Bubble tube constant head permeameter. 
Figure B.3- Nuclear density gauge for compaction and moisture content testing. 
Figure B.4- Lightweight deflectometer 
viii 
 
Figure B.5- Geogauge used in field testing (After Humboldt Mfg.) 
Figure B.6- Field hydraulic conductivity test on GAB. (ASTM D6391--Stage 1) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
1. INTRODUCTION 
 
Millions of tons of graded aggregate base (GAB) materials are used in construction of 
highway base layers in Maryland due to their satisfactory mechanical properties.  The 
fines content of a GAB material is highly variable and is often related to crushing 
process, stockpiling in the quarry, transportation and lying at the site.  The crushing of the 
stone at the quarry generally does not decrease the mechanical strength and stiffness of 
the material delivered to the site. However, Maryland State Highway Administration 
(SHA) is experiencing difficulties in achieving proper drainage through the base layers 
due to occasionally high fines content of the delivered GABs.  Presence of excessive 
water in pavement systems is one of the main reasons that cause pavement distress which 
decreases the service life of the pavement structures significantly (NCHRP 1997).  The 
relatively impervious base-course materials may shorten the service life of highways and 
increase the deterioration of the upper surface (asphalt layer) of pavements (NCHRP 
1997). 
Drainage in pavements can only be achieved with a properly designed and 
constructed system that consists of all essential drainage components and a base layer 
with adequate drainability and sufficient structural stability. The presence of free 
moisture in pavement layers has been found responsible for many premature failures 
observed in both flexible and rigid pavements (Abhijit et.al 2011).  Diefenderfer et.al. 
(2001) presents six adverse effects of excess water in pavement life: reduction in shear 
strength of the unbound material, increase in differential swelling of expansive subgrade 
soil, movement of fines in base and subbase layers, frost-heave and thaw weakening, 
cracking in rigid pavements, and stripping of asphalt in flexible pavement. Erlingsson et 
 2 
 
al. (2009) used heavy vehicle simulator to show that the rate of rutting depth increased in 
all layers of flexible pavement structure when the groundwater table was raised. Dawson  
(2009) has also shown that poor drainage pose significant adverse effects on the 
condition of roadways. Free moisture in the pavement sublayers largely occurs due to 
infiltration of rainwater and melted snow through pavement surface joints or cracks. To 
mitigate the moisture-induced distresses, it is imperative to drain free moisture out of 
pavement structures as quickly as possible via a good drainage system. Although the 
performance of a subsurface drainage system depends on all of its individual 
components, the hydraulic conductivity of a highway base layer can be critical for its 
adequate drainage (NCHRP 1997). Several factors, including GAB physical and chemical 
properties of aggregates, geometry of pavements, climatic conditions, and pavement 
surface conditions, affect the minimum hydraulic conductivity of a highway base layer or 
the time to achieve a certain percentage of drainage in the pavement structure 
(Casagrande and Shanon, 1952). 
In addition to high a quality drainage system, highway base layers should also 
have satisfactory mechanical properties such as high resistance to permanent (plastic) 
deformation under normal traffic loading. Therefore, it is imperative to consider 
structural stability in the optimization of highway base materials. Traditionally, the 
California bearing ratio (CBR) test has been used to quantify the structural stability of 
highway base materials due to its simplicity; however, it does not represent the stiffness 
of soils at low strains. Accordingly, SHA is no longer evaluating the pavement 
performance solely based on CBR test results.  The resilient modulus is arguably superior 
to static tests, such as CBR, due to its capability of characterizing the response of 
 3 
 
pavement material under repeated loading that simulates traffic loading conditions 
(AASHTO T 307).  The resilient modulus test provides an essential input parameter for 
the pavement design and the permanent deformation test provides information on the 
rutting potential of a pavement material in field conditions.  
There is no agreement on the minimum value of hydraulic conductivity or the 
time to achieve a given percentage of drainage; however, hydraulic conductivity and the 
appropriate drainage time are the indicators of pavement service life. Similarly, the 
minimum structural stability required for a permeable aggregate base is not well 
established in the previous studies and design guidelines. Therefore, there is a need to 
identify a range of gradation for highway base materials that can provide a better 
characterization of structural stability along with the high quality drainage in highway 
base layers.   
To respond to this need, a battery of tests was conducted on graded aggregate 
base (GAB) course materials, in the laboratory as well as in the field. Recycled concrete 
aggregates (RCAs) and their mixtures with select GABs were also included in the 
laboratory testing program since beneficial reuse of RCA brings economical advantages 
due to a decrease in its disposal associated with clogging of landfill leachate collection 
systems.  California bearing ratio (CBR), resilient modulus, permanent deformation and 
hydraulic conductivity tests were conducted to investigate the engineering properties of 
GAB, RCA and their mixtures, and to study the effect of curing time on RCA.  The effect 
of winter conditions were also evaluated by performing resilient modulus tests on the 
RCA specimens after a series of freeze-thaw cycles. 
 
 4 
 
2. MATERIALS 
 
 
The graded aggregate base materials (GABs) included in the current testing program are 
commonly used as highway base materials by the Maryland State Highway 
Administration (SHA).  GAB contains coarse and fine aggregate particles as well as fines 
(clay and silt). Generally, the ratio between coarse and fine aggregate particles varies 
between 1:1 and 7:3. Seven GAB materials were collected from different quarries in 
Maryland and tested in the laboratory. GABs used in the current study were named on the 
locations of the quarries they were collected from: Chantilly, Havre de Grace, Keystone, 
Churchville, Texas, Bladensburg, and Rockville.  The petrographic data shows that all 
GAB materials used in this study had different mineralogy (Table 1).  All GAB materials 
met the SHA and AASHTO M-147 specifications and were classified as high quality 
base materials (A-1-a (0) according to AASHTO Soil Classification System. 
The gradations of all selected GAB materials were within the tolerance limits of 
SHA specifications except few fractions of Texas quarry (Figure 1). The index properties 
of GAB materials are shown in Table 1. All GAB materials were non-plastic and their as-
 received fines content ranged from 6.9% to10%. According to SHA specifications, the 
fines content of the GAB materials used in highway base layers must be less than 8% 
(SHA 2012).  The absorption of fine and coarse aggregates of GAB materials varied 
between 0.89 and 5.33% and 0.4 and 0.79% respectively (Table 1). The Los Angeles 
abrasion values of all GABs were below 30% except Texas GAB.  Petrographic and 
mineralogical nature (marble, high CaCO3 content) of Texas GAB may have caused the 
relatively higher loss during the abrasion tests as marble stone tends to be easily crushed 
 5 
 
under impact loading. The micro deval values of GAB materials were 7.5- 24.8%.   The 
Rockville and Texas quarries yielded high micro deval values (21.9 and 24.8%, 
respectively), indicating that these GAB materials were not durable under moist 
conditions. Mineralogical natures and shapes of the Rockville and Texas GAB particles 
could be the reason for high micro deval values. The percentage loss in sodium sulfate 
test for the GAB materials ranged from 0.6 to 2.2%, meaning that all GAB materials had 
good resistance against freezing and thawing process.  
Two Maryland RCA materials, named A and B, were also included in laboratory 
testing program.  RCAs were generated from the demolition of concrete structures and 
stockpiled in Plants A and B located in Maryland. SHA does not have any highway 
construction specification that deals with the usage of RCAs.  The fines content of 
materials A and B were measured as 6 and 9%, respectively, and grain size distribution 
curves of both materials were within the SHA GAB limits (Figure 1). The absorption 
values of both RCAs were 4.2%; however, the Los Angeles abrasion of A exceeded the 
specification limit of 50%.  The percent losses based on sodium sulfate tests were 15.7 
and 14.3% for A and B, respectively, and exceeded the SHA specification limit of 12%, 
which could be due a reaction of sodium sulfate with cement contents present in material.  
The physical and chemical properties of the two RCA materials are summarized in Table 
1. 
  
 6 
 
3.  METHODS: 
3.1 Laboratory Geomechanical Tests 
 
The California bearing ratio (CBR) test is a penetration test for evaluation of the 
mechanical behavior of road base and subbase course layers. The CBR tests were 
performed on Rockville, Bladensburg, Churchville, and Texas GAB materials, the two 
RCA materials, and their mixtures with Bladensburg and Rockville GABs. Bladensburg 
and Rockville GABs were blended with two RCAs at 75:25, 50:50, and 25:75 ratios by 
weight.  Two types of compaction methods were utilized to observe the effect of 
compaction on CBR: impact Modified Proctor compaction (ASTM D1557) and vibratory 
compaction (ASTM D7382).  The specimens for vibratory compaction was prepared in 
three equal layers using a vibration frequency of 55 Hz for 60?5 seconds per layer.  A 
BOSCH 11248 EVS model vibratory hammer was used.  All specimens were compacted 
at their optimum moisture contents (OMC).  Table 1 provides the optimum moisture 
contents (OMCs) and maximum dry unit weights ( dm) of the GAB and RCA materials. 
All CBR tests were conducted by following the methods outlined in AASHTO T-193 and 
ASTM D 1883.  The specimens were un-soaked and the tests were performed at a strain 
rate of 1.27 mm/min.  
Resilient modulus test provides the stiffness of a soil under a confining stress and 
a repeated axial load.  The procedure outlined in AASHTO T 307-99, a protocol for 
testing of highway base and subbase materials, was followed for resilient modulus tests. 
All specimens were compacted by vibratory compactor in split mold of 152 mm in 
diameter and 305 mm in height, following the sugestions of Cetin et al. (2010).  The 
photo of the resilient modulus test equipment is shown in Figure B-1. Resilient modulus 
 7 
 
tests were performed on GAB, RCA and mixtures of GAB and RCA prepared at the same 
ratios of those tested for CBR. Each sample was compacted in six layers at their optimum 
moisture contents (OMC) and maximum dry densities using a vibratory compactor 
(ASTM D73820).  RCA specimens were removed from the molds after compaction, 
sealed in plastic wrap, and cured at 100% relative humidity and controlled temperature 
(21
 ?
 2 OC) for 1, 7 and 28 days before testing.  In order to evaluate the effect of moisture 
contents on resilient modulus (MR), specimens of all GABs were prepared and tested at 
2% below and 2% above the OMC. Resilient modulus tests were also performed on GAB 
samples collected from the construction sites. The field samples were collected from the 
locations where geogauge, nuclear density gauge, and light weight deflectometer (LWD) 
tests were conducted. The laboratory resilient modulus tests were conducted on field-
 retrieved GAB samples prepared at their field gradations, moisture contents, and 
compaction levels.  
To determine the climate effects on the mechanical properties of RCAs, 
specimens were prepared at OMC and maximum dry density in split molds and cured for 
28 days before subjecting them to 1, 4, 8, 16, and 20 cycles of freezing and thawing (F-T) 
per ASTM D6035. Each F-T cycle consisted of exposing each specimen to -19?C for 24 
hours, followed by room temperature (~20?C) for another 24 hours. The effect of F-T 
cycling on the engineering properties of recycled materials was determined by 
conducting resilient modulus tests after selection of the corresponding F-T cycles. 
Duplicate specimens were tested for most of the resilient modulus tests as quality control.  
A Geocomp LoadTrac-II loading frame and associated hydraulic power unit 
system was used to load the specimens.  The specimens were subjected to conditioning 
 8 
 
before the actual test loading under the confining and axial stress of 103 kPa for 500 
repetitions. Confining stress was kept between 20.7 and 138 kPa during loading stages, 
and the deviator stress was increased from 20.7 kPa to 276 kPa and applied 100 
repetitions at each step. The detailed information about the load sequences are provided 
in Table A-1. The loading sequence, confining pressure, and data acquisition were 
controlled by a personal computer equipped with RM 5.0 software.  Deformation data 
were measured with external linear variable displacement transducers (LVDTs) that had a 
measurement range of 0 to 50.8 mm.  
Resilient moduli from the last five cycles of each test sequence were averaged to 
obtain resilient modulus for each load sequence.  This nonlinear behavior of unbound 
granular material was defined in this study using the model developed by Witczak and 
Uzan (1988) recommended the following formulation 
32
 3
 1
 k
 a
 d
 k
 a
 aR
 pp
 pkM ??
 ?
 ?
 ??
 ?
 ?
 ??
 ?
 ?
 ??
 ?
 ?
 ?
 ??                                    (1) 
where MR is resilient modulus, k1, k2, and k3 are constants,  is isotropic confining 
pressure, and d is the deviator stress, pa is atmospheric pressure. A summary resilient 
modulus (SMR) was computed at a bulk stress of 208 kPa, following the guidelines 
provided in NCHRP 1-28A. With few exceptions, high R2 values (R2 >0.9) were obtained 
from regression analyses performed on the model. 
AASHTO T-307 test guidelines were followed to run the permanent deformation 
tests. During the permanent deformation test, same preconditioning load sequence of 
resilient modulus tests was followed.  After the preconditioning stage, the specimens 
were subjected to 10,000 load repetitions under 103.4 kPa confining pressure and 206.8 
kPa deviator stresses. Permanent deformation tests were performed until either 10,000 
 9 
 
load repetitions were completed or the permanent deformation of the tested specimen was 
exceeded the original length of the specimen by 5%.  A series of laboratory tests were 
also performed to study the effect of  moisture content (OMC, OMC-2%, and OMC+2%) 
on permanent deformation. 
3.2 Laboratory Hydraulic Conductivity Tests. 
 
Hydraulic conductivities of the different GAB materials were determined using a rigid-
 wall permeameter that was specifically developed for testing of asphalt and GAB 
specimens (Kutay et al. 2007).   The GAB specimens were compacted in the mold having 
dimension of 203 mm diameter and 203 mm height by using a vibratory compactor in 
four to six equal layers. The test set-up allows application of a wide a range of hydraulic 
gradients and accommodates high flow rates that are associated with testing of permeable 
specimens, and significantly minimizes sidewall leakage. The unique design also 
eliminates the use of valves, fittings and smaller diameter tubing, all which contributes to 
head losses that interfere with the test measurements, yet follows all recommendations in 
ASTM D2434 (Figure B-2).   
The permeameter was placed in a bath to maintain constant tail water elevation.  
The tub rim was located a few millimeters above the specimen top. As the water flows 
out of the reservoir tube through the specimen, air bubbles emerge from the bottom of the 
bubble tube. The constant total head difference through the specimen (H) was the height 
difference between the bottom of the bubble tube and the top of the water bath, and was 
used to calculate the hydraulic gradient (i).  The total flow rate through the specimen (Q) 
was determined by noting the water elevation drop in the reservoir tube and multiplying it 
 10 
 
with the inner area of the reservoir tube (A). Finally, the vertical hydraulic conductivities 
(k) were calculated using Darcy?s law.  
3.3 Field Tests. 
 
A series of geogauge, nuclear gauge and light weight deflectometer (LWD) 
measurements were conducted on the highway test sections constructed with five 
different GABs. The construction sites were located at Inter County connector (ICC) 
(Rockville GAB), I-695 (Texas GAB), I-295 (Havre-de Grace GAB), MD 725 
(Bladensburg GAB), MD 231 (Chantilly GAB).  Samples were collected from each test 
site following the procedures outlined in AASHTO T-2.  The field-retrieved samples 
were transported to the laboratory and subjected to resilient modulus and hydraulic 
conductivity tests to compare their physical and mechanical properties with those 
collected from the quarries.  The grain size distribution curves of the field-retrieved 
samples are shown in Figure 2.  All gradations lie within the SHA upper and lower 
gradation limits, except two samples of the Texas GAB material, indicating that the test 
sections were generally built by conforming the SHA guidelines. 
3.3.1 Light Weight Deflectometer 
 
Light weight deflectometer (LWD) is designed to determine the surface modulus, a 
response of the underlying structure in terms of a transient deflection to the dynamic 
stress applied through a circular bearing plate. Test locations at the construction site were 
selected on the basis of geometry of the road.  A series of density and moisture content 
measurements were performed via nuclear density gauge (Figure B-3) at the same 
locations where LWD tests were executed. 
 11 
 
The test procedure outlined in ASTM E2583 was followed to conduct the LWD 
tests. The base plate of the LWD equipment was placed on a flat and smooth surface, and 
dynamic load on the ground was applied by dropping 10kg load from 0.5 m height. These 
measurements were at least three times at the same location and an average of the 
measuremements was recorded as the modulus value of the tested location.  This 
deflection response is a composite response from the underlying structure within the zone 
of influence, which is dictated by a combination of the plate diameter, applied dynamic 
load and characteristics of the underlying materials. The zone of influence for the test 
may extend to a depth equal to 1-1.5 times the plate diameter, i.e. testing undertaken with 
a 300 mm plate is likely to have a zone of influence between 300 and 450 mm depth. The 
following model was used to calculate the LWD-based modulus:  
 
? ?
 o
 oo d
 afE ????? ?? 21
                                   (2) 
 
where Eo is surface modulus (Mpa), f is plate rigidity factor, v is Poisson?s ratio (~0.35), 
?o is maximum contact stress (MPa), a is plate radius (mm), and do is maximum 
deflection (mm). 
LWD used in the study had a fixed-drop height, and deflection was measured via 
an accelerometer mounted rigidly within the middle of the bearing plate to measure. 
According to ASTM E2583, the initial 1-3 drops were considered to provide a ?seating 
pressure? to ensure good contact, and further drops were used to determine the surface 
modulus. A photo of the equipment is given in Figure B-4. 
 12 
 
3.3.2 Geogauge. 
 
Geogauge was used to determine the stiffness and young?s modulus of GAB materials at 
the same locations where LWD and nuclear gauge tests were performed (Figure B-5).  
Geogauge was placed on the ground, and slightly rotated to achieve sufficient contact 
between foot of the geogauge and ground. On hard or rough surfaces, seating of the foot 
was assisted by the use of less than 10 mm (1/4") thickness of moist sand. 
Geogauge is a hand-portable instrument that provides stiffness and material 
modulus (NCHRP 10-65). The device measures the force imparted to the soil and records 
the resulting surface deflection as a function of frequency. Stiffness, force over 
deflection, follows directly from the impedance. Geogauge imparts very small 
displacements to the ground (< 1.27 x 10-6 m or <5 x 10-5 in) at 25 steady state 
frequencies between 100 and 196 Hz. Stiffness is determined at each frequency and the 
average stiffness for the 25 frequencies is displayed in N/m. The entire process takes 
about one minute. At these low frequencies, the impedance at the surface is stiffness-
 controlled and is proportional to the shear modulus of the soil.  The stiffness, K (N/m), is 
calculated using the following equation: 
)1(
 77.1
 2?? ?
 ??
 REP
 K
                                             (3) 
where P (N) is load, delta  (m) is deflection, R (m) is radius of the contact ring, E 
(N/m2) is shear modulus and  is Poisson ratio. 
3.3.3 Field Hydraulic Conductivity Tests. 
  
A series of borehole hydraulic conductivity tests were conducted at the construction sites, 
following the procedure outlined in ASTM D6391 (Figure B-6).  The first stage of the 
 13 
 
test method was employed as it provides the vertical hydraulic conductivity. Bentonide 
was used as a sealant around the borehole, and tests were performed on the basis of 
falling head method.  Furthermore, GAB samples were collected from each test site and 
compacted to field density and water content upon transporting to laboratory.  A series of 
laboratory hydraulic conductivity tests were performed on the field-retrieved samples by 
following the procedures outlined in Section 3.2.  
  
 14 
 
4.  RESULTS AND ANALYSIS 
4.1 CBR Tests.  
Table 2 shows the CBR results for GAB materials. The CBR of Bladensburg GAB was 
the highest (218) among others while the Rockville GAB resulted in the lowest CBR 
(68). The reason of such variation in CBR values of GAB materials could be different 
gradations, packing arrangement of particles and fines content. It is well-known that the 
structural stability of an unbound aggregate is affected by its particle size distribution 
(gradation), particle shape, packing arrangements, and angularity of the coarse particles 
(White et al. 2002). The CBRs of all GABs prepared with impact compaction method are 
significantly lower than those prepared with the vibratory compaction. Fines content 
increased due to crushing of the coarse aggregate during the impact compaction process 
as shown in Figure A-1. Siswosoebrotho et al. (2005) showed that coarse materials 
contained more than 4% fines content decreased the CBR value since excessive fines 
caused reduction in interlocking between the angular aggregates, which may have 
influenced the strength of the coarse material.  The data by Bennert and Maher (2005) 
also revealed that an increase in fines content decreases the CBR values significantly, 
consistent with the findings obtained in the current study. Therefore, all GABs were 
compacted by a vibratory hammer before performing resilient modulus, permamnent 
deformation, and hydraulic conductivity tests.   
Table 3a shows that the RCA specimens cured for 1 day resulted in lower CBR 
than those subjected to 7 day-curing.  Poon et al. (2006) stated that unhydrated cement 
content retained within the adhered mortar was the cause of self-cementing in RCA used 
as unbound base. Table 3b presents the results of CBR tests performed on mixtures of 
 15 
 
RCA and GABs.  Blandensburg-based mixtures resulted in higher CBRs but a consistent 
trend cannot be observed with CBR value and percent RCA addition.  
4.2. Resilient Modulus Tests 
 
Average SMR of GAB materials collected from construction sites and quarry locations 
are shown in the Table 2. The difference in SMR of quarry and field-collected samples 
may be due to a change in gradation, moisture content, fines content, and unit weight.  
Several studies suggest that the resilient modulus generally increases with an increase in 
density of the tested material (Robinson 1974, Rada and Witczak 1981, Kolisoja 1997).  
The number of contacts per particle increases significantly with increased density 
resulting from additional compaction of the particulate system. This, in turn, decreases 
the average contact stress corresponding to a certain external load. Hence, the 
deformation in particle contacts decreases and the resilient modulus increases (Kolisoja, 
1997). 
Figure 3 shows that GAB resilient modulus increases considerably with an 
increase in bulk stress, consistent with the findings of previous studies (Hicks 1970, 
Smith and Nair 1973, Uzan 1985, and Sweere 1990). Table 2 and Figure 4 show that the 
resilient moduli of the GAB materials prepared at OMC-2% were generally higher than 
the moduli of those prepared at OMC+2% except Texas GAB.  These findings were 
consistent with the previous studies. For instance, Smith and Nair (1973) and Vuong 
(1992) indicated that the resilient response of dry and partially saturated granular 
materials was high, but as complete saturation was approached, the resilient behavior of 
these materials had been affected significantly. Past research also revealed that the 
resilient moduli of granular materials were highly dependent on the moisture levels and 
 16 
 
tended to decrease near saturation (Haynes and Yoder 1963, Hicks and Monismith 1971). 
Furthermore, Dawson et al. (1996) studied a range of well-graded unbound aggregates 
and determined that, below the optimum moisture content, stiffness tended to increase 
apparently due to development of suction. Beyond the optimum moisture content, as the 
material became more saturated and excess pore water pressure was developed, the trend 
was shifted and stiffness started to decline rapidly. 
An exception to the observed trend in Figure 4 was with the Texas GAB material, 
which experienced a maximum SMR when compacted at its optimum moisture content. 
Thom and Brown (1987) observed a similar behavior in testing of select aggregates and 
attributed it to lubricating effect of moisture on particles that would decrease the 
deformation in the aggregate assembly and yield a resilient modulus increase. 
Figure 5 shows the effect of fines content and gravel/sand ratio on the resilient 
moduli of Rockville GAB material, respectively. SMR increased 2-2.5 times with an 
increase in fines content from 2 to 8% by weight, and then gradually decreased with 
further addition of fines. A nearly bell-shaped relationship can be observed when the 
gravel-to-sand ratio is plotted against SMr (Figure 5b).  SMr increases nearly 1.4 times 
with a change in G/S ratio from 1.5 to 1.7, and reaches to its maximum value at the 
optimum G/S value of ~1.7.  Further decrease in sand fractions makes the material 
unstable depending on the gravel size distribution.  Similar observations were made by 
Xiao et al. (2009) on testing of three aggregates with different petrographic natures. 
Previous studies reported that the resilient modulus of granular materials generally tended 
to decrease with an increase in fines content (Thom and Brown 1987; Kamal et al. 1993). 
Jorenby and Hicks (1986) also showed that initially an increase in fines content provided 
 17 
 
higher stiffness in granular materials and then a considerable reduction in stiffness was 
observed as more fines content were added to a crushed aggregate.  
The trends in Figure 5 can be explained by the influence of fines addition packing 
of the particles in a soil matrix.  Figure 6 shows the hypothetical packing arrangements of 
coarse particles with varying fine contents in a soil matrix. Coarse aggregates can be 
interlocked with each other that yield lower density and large voids due to lack of fines, 
as shown in Figure 6a.  This kind of soil matrix is referred to as gap graded gradation, 
and brings several advantages. The matrix provides good drainage and is less susceptible 
to frost and heave processes. Xiao et al. (2012) claimed that the GABs at this state may 
develop an unstable permanent deformation behavior.   The soil matrix shown in Figure 
6b is classified as dense-graded in which most of the voids between the aggregates are 
filled by fines but coarse particles are still in contact with each other. The grain-to-grain 
contact and void filling by fines are the possible reasons for the strength gain in this state. 
This soil matrix provides higher density which yields higher stiffness yet decreases the 
hydraulic conductivity. On the other hand, the compaction of such a soil matrix is 
moderately difficult. There is no grain-to-grain contact of aggregates in a soil matrix 
shown in Figure 6c. It has reasonably low density and hydraulic conductivity and the 
coarse particles are floating in the fine particles. The compaction of this kind of soil 
matrix is easier; however, its stability can easily be affected by adverse water conditions.  
The initial improvement in stiffness observed in the current study was attributed 
to increased contacts during pore filling as explained (Figure 6b). Addition of excessive 
fines gradually displaced the coarse particles in the soil matrix and caused stiffness to 
decrease (Figure 6c). The initial increase in resilient modulus with an increase in fines 
 18 
 
content was due to packing of aggregates which decreased the recoverable strain of the 
material and resulted in stiff material (Figure 5b).    
Resilient modulus tests were also performed on RCAs and mixtures prepared at 
varying RCA-to-GAB ratios. It can be seen from Figure 7 and Table 3b that 100%RCA 
and 100%GAB provide relatively higher MR values as compared to their mixtures, with 
few exceptions.  Similar observations were made by Kazmee et al. (2012) who attributed 
this behavior to poor packing of particles and change in gradation parameters. Table 3a 
indicates that the SMR of RCAs tend to increase with an increase in freezing and thawing 
cycles. The trends are reported by Bozyurt et.al (2011).  The stiffness increase in the 
current study is attributed to the continuation of hydration (cementation) reactions in 
RCA during the freeeze-thaw cycles. 
4.3. Permanent Deformation Tests 
 
Granular materials exhibit permanent deformation if they are subjected to repetitive 
loading for extended periods of time. The permanent deformation values are strongly 
dependent on rigidity, shear stress and load capacity of the granular materials. Use of the 
resilient modulus by itself is not sufficient to fully characterize the mechanical behavior 
of a pavement structure and should be coupled with permanent deformation tests 
(Khogali and Mohammad, 2004).   
Figure 8 shows the variation of cumulative permanent axial strain (plastic strain) 
with applied number of load repetitions. To model the relationship between the applied 
number of load repetitions and plastic strain, a power model was used: 
baN
 P
 ??
      (4) 
 19 
 
 
where a, b are fitting parameters; P is the cumulative permanent axial strain and N is the 
number load repetitions.  The permanent deformation (i.e., plastic strain) depends on the 
packing arrangement of particles, grain size distribution, and particle contact area. Table 
2 shows the plastic strain of all GAB materials used in the current study.  Keystone GAB 
had the maximum plastic strain (0.09%) while Texas GAB had the minimum plastic 
strain (0.03%) among all GAB materials after 10,000 repeated cycles of loading, which 
may be attributed to the gravel contents of the two GABs.  The Keystone GAB matrix 
has higher gravel content (56% versus 46%) and thus includes more voids (Figure 6a).  
Such a matrix, due to lack of good amount of fines and sand,  include gravel-to-gravel 
contact only and may experience more deformation during repeated loading (Xiao et al. 
2012).   
The data in Figure 9 suggest that GABs compacted at wet side of optimum are 
more susceptible to structural rutting. Khogali et al. (2004) also observed 2-3 times 
increase in permanent deformations of roadway bases due to fluctuations in groundwater 
table. Uthus (2007) claimed that dry density, degree of saturation, and stress level seemed 
to be the key parameters for influencing the permanent deformation behavior, along with 
mineralogy, fines content and grain size distributions of the granular materials. Increase 
in moisture content from OMC-2% to OMC for all GABs yielded approximately the 
same amount of plastic strain under long term of repetitive loads (Figure 9). The addition 
of moisture above OMC caused pore water pressure increase under repeated loading and 
resulted in excessive deformations.  
 20 
 
As shown in Figure10, the permanent deformation of GAB increases upon mixing 
with RCA, suggesting higher likelihood of rutting of a pavement system built with 
GAB/RCA blends.  Similar observations were made by Kazmee et al (2011). It is also 
noted that the plastic strain in individual GAB and RCA materials is less than their 
mixtures. This could be due to poor packing arrangement of particles when these two 
materials are mixed. 
4.4. Laboratory Hydraulic Conductivity Tests 
 
Table 4 summarizes the hydraulic conductivities of seven GABs tested in the laboratory 
and in-situ.  In-situ hydraulic conductivities are only 0.76-1.64 and 0.6-1.48 times higher 
than the laboratory-measured hydraulic conductivities of the samples collected from the 
quarries and field test locations, respectively.  The difference is less than an order of 
magnitude, suggesting that the laboratory and field hydraulic conductivities are 
comparable. 
In order to study the effect of fines content on hydraulic conductivity, gradations 
of Rockville and Texas GAB materials were adjusted by following two different 
approaches.   First, gradation was adjusted between the US. No  #4 and #30 sieves and 
rest of the fractions were kept constant (Figure 11a). Figure 12a  shows that such an 
adjustment does not significantly alter hydraulic conductivity of Rockville GAB. These 
results confirm the commonly observed trend that coarser portion of sand in GAB (e.g., 
between the U.S. #4 and #30 sieves)  do not have a significant effect on hydraulic 
conductivity and flow is mainly controlled by smaller particles in the gradation (Cote and 
konrad 2003). Therefore, at the second stage, adjustements were made for the fractions 
between the 9.5-mm (3/8-inch) and 12.7-mm (1/2-inch), and 19-mm (3/4-inch) sieves to 
 21 
 
represent a more widespread change in grain size distribution (Figures 11b and 11c).  The 
data in Table 5 reveal that Rockville and Texas GAB hydraulic conductivities reduced 5 
and 50 times, respectively, as a result of such fines content adjustment from 2 to 16% 
(Figures 12b and 12c). Similar observations were reported by Siswosoebrotho et.al 
(2005) during testing of unbound granular materials.  
Figure 13 shows that hydraulic conductivities of Rockville and Texas GAB 
materials increase up to 5 and 50 times, respectively, with nearly 14% increase in gravel 
content for both materials. Similar magnitudes of increase in hydraulic conductivity were 
observed when gravel-to-sand (G/S) ratio was varied between 1.45 and 1.85 and 0.82 and 
1.14 for Rockville and Texas GABs, respectively (Figure 14).  Analysis of the trends in  
Figure 14 shows that the soil matrix is porous and leads to higher hydraulic conductivities 
when G/S ratio >1.7 for Rockville and G/S>1.05 for Texas GABs.  It is believed that the 
minimum porosities are achieved at these gravel-to-sand ratios due to optimum packing 
of the GAB medium.  Xiao et al. (2012) also reported minimum porosity achievements at 
G/S=1.56-1.68 and G/S~1.5, respectively, for GABs with varying petrography.   
Figure 15 shows that hydraulic conductivites of Rockville and Texas GABs are 
increased up to 5 and 50 times, respectively, with increasing characeterictic grain sizes of 
the soil (i.e., D10, D30 , D50 and D60). The hydraulic conductivity seems to be more 
sensitive to the smaller grain sizes (D10 and D30) as compared to larges sizes (D50 and 
D60), consistent with the previous studies that finer sizes play a major role on hydraulic 
conductivity (FHWA 2005). 
  
 22 
 
4.5 Field Tests 
 
The in-situ stiffness and modulus values of the GAB materials were measured via light 
weight deflectometer (LWD) and geogauge, and the data are summarized in Table C1 of 
Appendix C. The field stiffness and moduli of the GAB materials are plotted against 
laboratory determined SMR in Figure 16. In order to determine the correlation between 
the laboratory resilient moduli and the moduli/stiffness obtained from geogauge and 
LWD, a paired t-test was conducted for statistical significance by determining whether 
the Pearson correlation coefficient between laboratory and field resilient 
modulus/stiffness is statistically different from zero.  For this statistical analysis, the t-
 statistic (t) was computed from the correlation coefficient (r) as: 
 
                                      
2
 1 2
 ?
 ?
 ?
 ?
 n
 r
 r
 t
 ?                    (5)                             
 
where  is the population correlation coefficient (assumed to be zero) and n is the 
number of degrees of freedom.  n was equal to 54, 5, and 5 for geogauge versus LWD 
test data, geogauge versus laboratory SMR data, and LWD versus laboratory SMR data. A 
comparison was made between t and the critical t (tcr) corresponding to a significance 
level .  If t > tcr, then the Pearson correlation coefficient was significantly different from 
zero and a significant relationship was assumed to exist between laboratory and field 
resilient modulus.  In this analysis,  was set to 0.05 (the commonly accepted 
significance level), which corresponds to tcr = 2.011 for geogauge versus LWD data and 
tcr = 3.182 for geogauge versus laboratory SMR, and LWD versus laboratory SMR data.    
 23 
 
Figure 16 shows that the correlation between geogauge and LWD is high 
(t=9.6>tcr=2.011). The coefficient of determination, R
 2, for the correlation between the 
data produced by the two field equipment was fair (R2>0.65).   The differences in induced 
stress and depth of influence of the applied load provided by LWD, and geogauge could 
be the possible reasons for the observed correlation. Previous studies showed that applied 
stress level was the factor that had the most significant impact on the resilient properties 
of granular materials (Kolisoja (1997). Significantly higher R2 values were observed for 
the correlations between the mean laboratory SMR and LWD or geogauge data (R
 2=0.83-
 0.97). In addition, t values that were obtained from statistical analyses (t>tcr= 3.182) 
indicate that reasonably good correlations exists between the geogauge and laboratory 
SMR as well as LWD and laboratory MR data at LWD bulk stress level. All regression 
lines were forced to pass through the zero intercept because of rationality of relations.   
Figure 17 shows the field and laboratory hydraulic conductivity test results.  The 
differences in the laboratory and field hydraulic conductivity values are negligible 
considering the anisotropy in the field.  The drainage qualities of the GAB materials 
tested in the laboratory and field can be considered as ?fair to good? according to the 
hydraulic conductivity range provided in AASHTO Guide (1993). 
  
 24 
 
5.  PRACTICAL IMPLICATIONS 
 
5.1 Highway Base Design  
 
Resilient modulus test results were used to estimate the thickness of the base layer in a 
pavement by following the procedures defined in the AASHTO Guide (1993). The 50 
million EASL value was assumed for this analysis. The overall standard deviation (So) 
and reliability (ZR) were assumed to be 0.35 and 95%, respectively.  Structural numbers 
(SN) were back-calculated using the following equation: 
 
 
  (6) 
where ?PSI  is design serviceability loss and MR is the roadbed material effective 
resilient modulus.  The values were selected as 1.9 and 34.5 Mpa, based on Huang 
(1993).  An asphalt layer thickness of 203.2 mm was selected.  The resilient modulus of 
asphalt was assumed to be 2965 MPa, which corresponded to a layer coefficient of a1 = 
0.44 according to AASHTO Guide (1993).  A resilient modulus of 103 MPa 
(corresponding to a structural coefficient of a3 = 0.08) and a thickness of 152.4 mm (D3) 
were assumed for the subbase layer. The laboratory-based SMR values vary between 120 
Mpa (17000 psi) and 210 Mpa (30500 psi) which correspond to a layer coefficient (a2) of 
0.08-0.14 according to AASHTO pavement design guidelines (1993). SMr of 206.84 
MPa and a2 of 0.12, the two values commonly used by SHA in absence of 
measuremements, fall within this range.  Finally, the base thicknesses were calculated 
using the following formula:                         
? ? ? ?
 ? ? ? ?
 ? ?
 ? ? 07.8log32.2
 110944.0
 5.12.4log
 20.01log36.9log 1019.5
 10
 10018 ???
 ??
 ??
 ??????? RR M
 SN
 PSI
 SNSZW
 25 
 
 
                         
m2a
 mD
 3
 a
 1
 D
 1
 aSN
 2
 D
 2
 33??
 ?
        (7) 
where m2 and m3 are drainage modification factors for base and subbase layer, 
respectively, and were chosen as 1.2, 1.0, 0.8, 0.6 for excellent, good, fair, and poor 
drainage conditions, respectively, within the pavement system (Huang 1993). D1, D2, and 
D3 are the layer thicknesses of asphalt layer, base layer, and subbase layer, respectively. 
It can be concluded from Table 6 that an increase in the base layer coefficient 
yields a decrease in required thickness of base layer while all other factors are kept 
constant. On the other hand, the decrease in drainage modification factor increases the 
required thickness of the base course. The effects of layer coefficient and drainage 
modification factor of GAB on the required design thickness are also reflected in Figure 
18.    
5.2. Effect of Hydraulic Conductivity on Highway Base Design 
 
Federal Highway Administration (FHWA) software DRIP (Drainage Requirement in 
Pavements) was used to evaluate the effect of hydraulic conductivity on drainage time 
and minimum required thickness of highway base layers. For the purpose of analysis, a 
typical cross section of highway having width (W) of 7.3 m (two lanes, each 3.65 m 
wide) was selected. The longitudinal slope (S) and cross slope (Sx) were considered as 
2%, and the resultant length of flow path (LR = W*[1+(S/Sx)
 2]1/2) was calculated.  
The largest source of water is the rain water that enters the pavement surface 
through cracks and joints in the surface. Two methods have been used to determine 
surface infiltration of water: the infiltration ratio method (Cedergren et al. 1973) and the 
 26 
 
crack infiltration method (Ridgeway 1976).  The infiltration ratio method is highly 
empirical and depends on both the infiltration ratio and rainfall rate.  The crack 
infiltration method, on the other hand, is based on the results of infiltration tests, and was 
preferred in the current analysis.  The equation to compute the infiltration rate for intact 
pavement is as follows: 
 
                         
p
 s
 cc
 ci k
 WC
 W
 W
 N
 Iq ??
 ?
 ?
 ?
 ?
 ?
 ??
                                     (8) 
 
where qi is rate of pavement infiltration (m
 3/day/m2), Ic is the crack infiltration rate, 
(m3/day/m), Nc is number of longitudinal cracks.  Ic and Nc were assumed as 0.22 
m3/day/m, and 3, respectively. The length of contributing transverse joints or cracks (Wc, 
m), the width of  base (W, m), and the spacing of contributing transverse joints or cracks 
(Cs, m) were 7.3 m, 7.92 m, and 7.3 m, respectively.  kp is pavement hydraulic 
conductivity (m/day) and a value of 0.051 m/day was assumed per Kutay et al. (2007).  
Two approaches were used to evaluate the drainage ability of the GAB layers: 
Depth-to-flow design approach and time-to-drain approach. In the first approach, the 
concept is that the steady flow capacity of base layer should be equal to or greater than 
the inflow of rainfall. Moulton (1980) developed an equation which presents that required  
base thickness (H) as a function of GAB hydraulic conductivity (k), slope (S) of highway, 
length of drainage (LR), and rate of pavement infiltration (qi).  The equations for depth-to- 
flow are:  
 27 
 
?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ??
 ?
 ?
 ?
 ??
 ?
 ?
 ?
 ??
 ?
 ?
 ?
 ??
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?? 2/
 4
 1tan
 41
 22
 ?
 Sk
 q
 S
 Sk
 q
 S
 RLk
 q
 H
 ii
 i              if   (S2 ? 4qi/k) < 0    
 kiqS
 S
 i
 ii
 kqSS
 kqSS
 RLk
 q
 H
 /422
 2
 2
 /4
 /4
 1
 ?
 ??
 ?
 ??
 ?
 ??
 ??
 ??             if   (S
 2 ? 4qi/k) > 0           
                      1
 1
 ?
 ?? RLk
 q
 H i
                        if (S2 ? 4qi/k) = 0         (11) 
 
where S and LR were assumed as 0.0283 and 11.21 m, respectively.  The highway 
geometry is beyond the scope of this work, thus a sensitivity analysis was conducted with 
respect to hydraulic conductivity (k) only. The laboratory quarry GAB hydraulic 
conductivities listed in Table 4 were used in the analysis.  The results shown in Figure 19 
indicates that the required  base thickness is more influenced from GAB permeabilities at 
k<0.1 cm/s.  
The second approach for design of the GAB layers includes a series of 
calculations for the time to drain 50% of the infiltrating water. The AASHTO pavement 
design guideline (1993) categorizes the base layer as excellent, good, fair, and poor based 
on time for 50% drainage. The following equations developed by Casagrande and 
Shannon (1952) and Barber and Sawyer (1952) and embedded in the DRIP software were 
used to calculate the time to drain a specified percentage of the infiltrating water: 
 
Casagrande and Shannon (1952) 
 
? ?? ? kH
 Ln
 SU
 USS
 S
 S
 S
 SS
 S
 t e
 2
 1
 11
 1
 1
 12
 113/1
 1
 122
 122
 ln
 1
 ln
 4.0
 2.1 ??
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ??
 ?
 ??
 ?
 ?
 ?
 ?
 ? ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ??
       if U > 0.5     
 28 
 
kH
 Ln
 S
 US
 SUS
 S
 t e
 2
 1
 12
 113/1
 1
 2
 ln2
 4.0
 2.1 ??
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ? ?
 ?
 ?
 ?
 ?
 ?
 ?
 ?
 ??
               if U ? 0.5     
(13) 
 
Barber and Sawyer (1952)  
 
? ?? ? kH
 Ln
 SU
 US
 S
 S
 SSt e
 2
 1
 11
 1
 1
 2
 11
 4.21
 2.1
 log15.1
 4.2
 1log48.05.0 ??
 ?
 ?
 ?
 ?
 ?
 ??
 ?
 ??
 ?
 ?
 ?
 ?
 ?
 ???
    if 0.5 ? U ? 1.0       
 
kH
 Ln
 S
 U
 SUSt e
 2
 1
 2
 1
 8.4
 1log48.0 ???
 ?
 ?
 ??
 ?
 ?
 ???
            if 0 ? U ? 0.5   
(15) 
 
where t is time (hours) for percent drainage, U, to be reached, S1 is dimensionless slope 
factor (= H/LS).  L and ne are width and effective porosity of the GAB layers and were 
taken as 7.3 m and 20- 70% of total porosity, respectively.   
A series of analysis was performed to gage the influence of U, H, and k on time-
 to-drain.  U and H varied between 0 and 98%, and 5 and 60 cm, respectively.  A unit 
weight of 1601.8 kg/m3 (100 pcf) and specific gravity of 2.70 for GAB were used. The 
water in the voids cannot be drained by gravity flow due to capillary action present in the 
soil matrix, thus, effective porosities were assumed to be 20-70% of the total porosities 
(Moulton 1980) and utilized in the analysis.   Four different gradations of Rockville and 
Texas GAB materials, with fines content of 2, 4, 8, and 14%, along with the lower and 
upper SHA gradation limits were used in the drainage calculations.  Rockville GAB 
hydraulic conductivities ranging from 6.4 x 10-3 to 3.12 x 10-2 cm/s that correspond to 2-
 16% fines content were utilized.  The corresponding Texas GAB hydraulic conductivities 
ranged from 3.67 x 10-3 to 7.32 x 10-5 cm/s.  
 29 
 
The results are shown in Figures 20-23. The time-to-drain does not change 
significantly up to 50% drainage, and an exponential increase in drainage time exists for 
50%<U<98%  (Figure 20,22). Moreover, when the base thickness of Rockville GAB was 
increased from 6 to 60 cm, 57% and 48% decreases in drainage time were observed based 
on Barber and Sawyer (1952) and Casagrande and Shannon (1952) methods, respectively 
(Figure 21).  The corresponding decreases in drainage time for Texas GAB were 45% 
and 54%, respectively (Figure 23). 
The driving factor for time-to-drain of a highway base is the GAB hydraulic 
conductivity. The required base layer thickness (Moulton method 1980) with respect to 
hydraulic conductivity values obtained at 2-14% fines are shown for Rockville GAB in 
Figure 24 and Table 7.  At a specific base thickness of 0.3m, the time-to-drain (at 
U=50%) increases three times with change of fines content from 2 to 14% (24h to 75h, 
changing the corresponding AASHTO drainage quality classification from Good to Fair 
(Table 7).  The corresponding increase in time-to-drain for Texas GAB was from 218 
hours to 3768 hours (Table 7). A 17 times increase in time-to-drain indicated that 
material was clogged by fines. It can also be seen from Figures 11 and 12 that TexasGAB 
stays out of the SHA gradation limits and experience unacceptable hydrayulic 
conductivities when the fines content was greater than 6%. 
 Figure 25 and 26 present the variation in time-to-drian with percent drainage and 
base thickness for all GABs.  The Churchville and Chantilly GAB materials took long 
time to drain as compared to others due to their relatively lower hydraulic conductivities 
(Table 4). The effect of hydraulic conductivity on drainage performance and required 
base thickness can clearly be seen in Figure 27.  The GAB materials with lower hydraulic 
 30 
 
conductivities yielded higher time-to-drain at U=50% and required large base thicknesses 
for construction. 
 5.3. Cost Calculations 
 
A simple cost analysis was performed on all GAB materials. The design thicknesses of 
the base layers (Table 6) were calculated using Equation  6 by assuming a pavement 
structure number (SN) of 5 based on 50 million EASL value, a layer coefficient (a1) of 
0.44 for the asphalt layer, and a layer coefficient (a3) of 0.08 for the subbase layer.  The 
layer coefficient of base layer (a2) was varied between 0.08 and 0.14 based on laboratory 
SMR of the GAB materials, and the drainage coefficient of both base and subbase (m1 and 
m2) were assumed to remain in a range of 0.6-1.2. The average unit price of the GAB 
material was considered to be $80/m3 ($10/ yd3 per 6-in lift thickness) following the 2013 
price index table issued by the Maryland SHA. The listed unit price of a GAB material 
includes material, hauling, transportation and laying costs only.  
Lane widths in the United States can range from 3 m (low volume roads) to 5 m 
(highway ramps) in width, and a typical design lane width of 3.65 m was selected for the 
cost analysis in the current study. A two-lane roadway was considered.  The cost analysis 
summarized in Table 6 indicates that the GAB cost decreases with increasing drainage 
modification factor or layer coefficient, and vice versa. It can be seen from Figure 28 that 
the cost decreases 62% with the increase of quality of drainage from poor to excellent or 
time-to-drain from 10 to 0.08 days. A 42% cost decrease is noticeable with a layer 
coefficient increase of 0.08 to 0.14. The construction cost of 1-km highway varies from 
$79,833 to $368,387, which indicates that proper selection of a highway base layer 
 31 
 
coefficient and drainage modification factor has a significant impact on the construction 
costs. 
  
 32 
 
6. CONCLUSIONS 
The structural stability and drainability of pavement structures depend on the mechanical 
and hydraulic characteristics of graded aggregate base (GAB) materials. A research study 
was conducted to evaluate the drainage and mechanical properties of GAB materials 
utilized in Maryland highways. In addition to seven GAB materials, two recycled 
concrete GAB materials and their selected mixtures were studied. The observations are 
summarized as follows: 
1) The GAB resilient modulus increased when fines content was varied between 2 
and 8% and, started decreasing with further fines addition. SMr was maximized 
when the fines content was ~8% and gravel-to-sand ratio was 1.6-1.7.  A 
minimum gravel content of 70% (>4.75 mm) by weight and a maximum fines 
content of 8% by weight should be specified for highway base construction with 
the GABs tested.  This can be controlled by avoiding segregation of the GABs 
and their proper mixing by pig mill at the construction site.  
2) The GAB resilient modulus generally decreased with moisture addition above 
OMC during compaction. SMr values at OMC-2% were higher than those at 
OMC, with few exceptions. On the other hand, the permanent deformations (i.e., a 
measure of structural rutting) were doubled with 2% increase in moisture contents 
from the OMCs; however, no significant change in permanent deformations 
occurred at OMC-2%.  The findings suggest that the field compaction mositure 
content should be as close to OMC as possible.  
3) The RCAs experienced  0.98 to 2.1 times increase in SMR with increasing freeze-
 thaw cycles due to ongoing hydration process during freezing and thawing.   The 
 33 
 
SMR of RCAs mixtures were lower than the ones for 100%RCA and 100%GAB 
materials, with few exceptions. Similarly, permanent deformations of the mixtures 
were generally higher than those obtained for pure GAB or RCA. 
4) An addition of 4-6% fines over the SHA specification limit of 8% resulted in 2-5 
times decrease in the laboratory-based GAB hydraulic conductivities and led to an 
increase in time for 50% completion of the drainage from the highway base (from 
50 hr to 75 hr). The required base thickness based on Moulton method (1979) was 
also increased 2.5 times as a result of reduction in GAB hydraulic conductivity.  
The laboratory and field hydraulic conductivities were generally comparable and 
ratio of the laboratory-to-field hydraulic conductivity was 0.6-3.5.  
5) The correlation between the mean laboratory and field stiffness/modulus values 
were fair to acceptable (R2=0.65 to 0.9); however, further research is required to 
improve the accuracy of correlation between the laboratory and field 
stiffness/modulus values. 
6) If percentage of fine materials is not carefully controlled during construction 
process, the base layer built with GAB materials may experience clogging which 
may, in turn, initiate the deterioration of the upper pavement layer (asphalt layer).  
Considering the hydraulic conductivity, resilient modulus and permanent 
deformation data of the current study, the fines content should be limited to 8% 
and gravel-to-sand ratio should be kept between 1.6 and 1.7.  
7) A simple cost analysis suggested that improper selection of layer coefficient and 
drainage modification factor for the base layer may lead to immature failure or 
uneconomical design. In cases where clogging of the base is of concern, a 
 34 
 
drainage analysis should be conducted in addition to geomechanical testing for 
cost-based selection of the layer coefficient.  
 
  
 35 
 
 
 
 
 
 
 
 
 
 
 
TABLES 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 36 
 
 
Table 1: Physical and chemical properties of the GAB and RCA materials 
LA MD SS SiO2 Al2O3 Fe2O3 CaO
 Imp Vib Imp Vib F C F (%) C (%) % % % % % % %
 Rockville 23.93 24.70 5.80 4.70 2.55 2.77 5.33 0.78 16.40 21.90 1.60 Meta Basalt 60.88 13.27 9.43 2.93
 Texas 23.89 24.29 4.20 4.10 2.70 2.79 1.75 0.40 53.04 24.76 0.53
 Carbonate 
Morble 
44.10 3.04 1.57 26.83
 Churchville 24.79 25.54 5.30 4.80 2.91 3.01 0.89 0.49 26.90 18.50 2.20 Gneiss 47.71 15.61 10.95 11.90
 Bladensburg 24.30 24.66 4.80 4.50 2.72 2.83 3.09 0.55 23.60 7.56 1.10
 Carbonate 
Dolomite
 50.73 12.93 10.92 10.67
 Chantilly 24.84 - 5.40 - 2.86 2.99 3.18 0.79 22.20 7.50 0.73 Basalt 38.39 9.48 7.18 5.80
 Havre De Grace 23.38 - 5.20 - 2.75 2.79 1.32 0.58 22.90 11.50 0.60
 Quartz/Feldi
 spat
 2.36 0.70 1.31 29.31
 keystone 23.03 - 4.70 - 2.60 2.68 2.63 0.51 25.20 12.20 2.02
 Carbonate-
 Siliceous 
11.90 1.95 0.85 31.67
 Flanigan 20.17 - 9.50 - 2.29 2.49 9.23 4.20 55.20 16.80 15.70 - 51.54 4.61 2.55 16.94
 Bishop 20.14 - 9.50 - 2.29 2.53 9.05 4.19 47.40 18.40 14.26 : 61.24 4.02 1.87 13.06
 ?d: maximum dry density, Imp: impact compactor, Vib: vibratory compactor, Gs: specific gravity, F: fine contents, C: Coarse contents,  LA: Loss angeles abrasion test , 
MD: Micro deval test , SS: loss in Sodium Sulfate test, PD: Petrographic Description.
 Physical Properties
 G
 AB
 Material
 chemical Properties
 PD
 RCG
 AB
 Gs Absorption?d (KN/m
 3
 ) OMC (%)
 37 
 
 
 
Table 2: CBR, SMR, power fitting parameters and plastic strain of the GAB materials 
GAB Material  
CBR SMR (Mpa) 
Mean power model fitting 
parameters (Standard 
deviation) 
Plastic 
strain 
 (%)  Impact 
compaction 
Vibratory 
compaction 
Field- 
retrieved 
Lab 
 OMC-2  
(%) 
Lab  
OMC 
(%) 
Lab 
OMC+2 
 (%) 
k1 (?) k2(?) k3(?) 
Rockville 58 68 148 193 180 151 
1025 
(216) 
0.88 
(.09) 
-0.22 
(0.07) 
0.035 
Bladensburg 121 218 135 130 114 119 
1121 
(606) 
0.91 
(0.33) 
-0.16 
(0.09) 
0.04 
Churchville 175 200 NA 140 87 NA 
679.9 
(278) 
1.07 
(0.25) 
-0.16 
(0.06) 
0.06 
Texas 85 150 162 176 257 148 
803.9 
(310) 
1.08 
(0.28) 
-0.15 
(0.13) 
0.03 
Chantilly NA NA 160 126 101 119 
663.5 
(183) 
0.96 
(0.27) 
-0.15 
(0.08) 
0.07 
Havre de 
Grace 
NA NA 222 130 120 124 
894.4 
(118) 
0.72 
(0.09) 
-0.11 
(0.09) 
0.045 
Keystone NA NA NA 147 134 87 
666.6 
(139) 
0.96 
(0.12) 
-0.10 
(0.02) 
0.09 
Notes: CBR: California bearing ratio, SMR plastic: plastic 
strain of specimens after 10,000 repeated load cycles. NA: Not analyzed.  
 
 
 
 
 
 
 
 38 
 
 
 
 
 
Table 3a. Effect of curing time and freeze-thaw cycles on CBR and SMR of the two 
RCAs. 
RCA 
CBR SMR (Mpa) Mean power model fitting 
parameters (Standard 
deviation) 
Curing 
Time 
Freezing and Thawing Cycles  
1 
day  
7 
days 
0 4 8 12 14 20 k1 k2 k3 
A 148 167 102 114 100 NA 140 215 
355.8 
(17.2) 
1.40 
(0.09) 
-0.18 
(0.09) 
B 114 131 122 122 123 126 NA 130 
493.3 
(35.2) 
1.18 
(0.07) 
-0.13 
(0.07) 
 
 
 
Table 3b. CBR, SMR and power model fitting parameters of RCA/GAB mixtures. 
RCA/GAB 
mixtures 
CBR 
SMR  
(MPa) 
Power model fitting 
parameters 
k1 k2 k3 
25A75BL 282 140 1495 0.80 -0.04 
50A50BL 319 150 543.5 1.30 -0.10 
75A25BL 301 260 492.3 1.29 -0.11 
25A75R 209 160 1430 0.82 -0.20 
50A50R 131 130 356.5 1.54 -0.17 
75A25R 154 280 478.1 1.45 -0.34 
25B75BL NA 340 452.3 1.36 -0.05 
50B50BL NA 280 1689 0.68 -0.08 
75B25BL NA 120 2313 0.53 -0.18 
25B75R 141 70 510.61 1.29 -0.18 
50B50R 194 120 450.63 1.27 -0.11 
75B25R 189 150 356.25 1.39 -0.21 
Notes: A: RCA from Plant A, B: RCA from Plant B, R: Rockville,  BL: Bladensburg, 
NA: Not analyzed 
 
 
 
 
 39 
 
 
 
 
 
 
Table 4: Mean hydraulic conductivity of GAB materials tested in the laboratory and in-
 situ. 
GAB Material  
k laboratory of 
quarry samples 
(cm/s) 
k laboratory of field-
 retrieved samples 
(cm/s) 
k in-situ 
Rockville 2.73 x 10-2 2.52 x 10-2 2.07 x 10-2 
Bladensburg 1.28 x 10-2 1.30 x 10-2 6.20 x 10-3 
Churchville 1.48 x 10-3 NA NA 
Texas 6.57 x 10-3 7.23 x 10-3 7.05 x 10-3 
Chantilly 5.66 x 10-4 6.30 x 10-4 9.30 x 10-4 
Havre de Grace 1.48 x 10-2 6.90 x 10-3 4.25 x 10-3 
Keystone 3.92 x 10-3 NA NA 
NA: Not analyzed 
  
 40 
 
 
Table 5: Effect of gradation parameters on SMR and hydraulic conductivity of GAB 
materials. 
 
  
FC D10 D30 D60 D50 Sand Gravel G/S SMR  k
 (%) mm mm mm mm (%) (%) Ratio MPa (cm/s)
 2 0.17 2.14 10.27 7.1 34.4 63.6 1.17 - 1.01 x 10-2
 6 0.22 1.7 10 6.5 34.4 59.6 1.27 - 2.56 x 10-2
 8 0.2 1.8 10 6.5 34.4 57.6 1.33 - 3.85 x 10-2
 10 0.15 1.7 10 6.5 34.4 55.6 1.38 - 3.40 x 10-2
 12 0.1 2 10 6.5 34.4 53.6 1.44 - 6.33 x 10-2
 14 0.1 2 10 6.5 34.4 51.6 1.56 - 3.83 x 10-2
 2 0.41 2.7 10.7 7.8 34.4 63.6 1.85 86.2 3.12 x 10-2
 4 0.24 2.2 10.4 7.1 34.4 61.6 1.79 148.2 1.92 x 10-2
 6 0.16 1.9 10 6.8 34.4 59.6 1.73 130.2 1.83 x 10-2
 8 0.12 1.7 9.3 6.2 34.4 57.6 1.67 179.8 1.04 x 10-2
 10 0.08 1.3 9 5.8 34.4 55.6 1.62 126.7 1.11 x 10-2
 12 0.05 1.15 8.2 5.3 34.4 53.6 1.56 146.5 1.14 x 10-2
 14 0.03 0.97 7.8 5.05 34.4 51.6 1.5 132.6 6.33 x 10-3
 16 0.02 0.82 7.1 4.3 34.4 49.6 1.44 - 6.40 x 10-3
 2 0.16 0.9 9 5.2 45.8 52.2 1.14 - 3.67 x 10-3
 4 0.13 0.75 8.2 4.9 45.8 50.2 1.1 - 2.71 x 10-3
 6 0.11 0.61 7.5 4 45.8 48.2 1.05 - 1.17 x 10-3
 8 0.09 0.5 7 3.2 45.8 46.2 1.01 - 4.22 x 10-4
 10 0.08 0.42 6 2.7 45.8 44.2 0.97 - 4.60 x 10-4
 12 0.06 0.35 5.4 2.3 45.8 42.2 0.92 - 6.44 x 10-5
 14 0.05 0.3 4.7 1.9 45.8 40.2 0.88 - 7.00 x 10-5
 16 0.04 0.25 4 1.6 45.8 38.2 0.83 - 7.32 x 10-5
 Rockville QG 7.6 0.12 1.8 10 6.5 34.4 58.03 1.69 190 2.73 x 10-2
  Texas QG 8.6 0.12 0.5 7 3.2 45.8 45.6 1 257 6.57 x 10-3
 SHA Spec. LL 0 0.4 3 12 9.5 36 64 1.78 - 3.19 x 10-2
 SHA.Spec UL 8 0.09 0.85 6 3.2 48 44 0.92 - 3.84 x 10-3
 FC: Fines contents, D10,D30,D50,D60 : Diameter of particles @ 10,30,50,60 passing percentage finer respectively, G/S: Gravel to 
Sand ratio, SMR: Summary Resilient Modulus, k : hydraulic conductivity, QG: quarry gradation, LL: Lower limit of SHA 
specified Gradation, UL: Upper Limit of SHA specified gradation
 Material
 Rockville Quarry 
Change in Fines 
content is adjusted in 
sand portion of 
gradation
 Rockville Quarry 
Change in fines 
content is adjusted in 
coarse portion of 
gradation
 Texas Quarry, Change 
in fines content is 
adjusted in coarse 
portion of gradation
 41 
 
 
Table 6. Effect of change in layer coefficient and drainage modification factor on the 
required base thickness in pavement design.  
a2 m3 =  m2 D2 (cm) Cost /km ($) 
0.08 
0.6 63.1 368,387 
0.8 43.5 254,040 
1 31.8 185,420 
1.2 23.9 139,693 
0.10 
0.6 50.5 294,686 
0.8 34.8 203,232 
1 25.4 148,336 
1.2 19.1 111,719 
0.12 
0.6 42.1 245,572 
0.8 29.0 169,360 
1 21.2 123,633 
1.2 16.0 93,148 
0.14 
0.6 36.0 210,474 
0.8 24.9 145,182 
1 18.1 105,938 
1.2 13.7 79,833 
Notes: a2: base layer coefficient, m2 :  base layer drainage modification factor, m3: 
subbase layer drainage modification factor, D2: required base thickness 
  
 42 
 
 
 
 
 
Table 7. Effect of fines content and GAB type on required base thickness and time to 
drain. 
Material Gradation 
Hydraulic 
 conductivity 
 (cm/s) 
Time to 
drain at 
U=50% 
(hours) 
AASHTO 
classification for  
quality of 
drainage 
R
 o
 ck
 v
 il
 le
  
2% FC 3.12 x 10-2 24 Good 
4% FC 1.92 x 10-2 36 Fair 
8% FC 1.04 x 10-2 52 Fair 
14% FC 6.33 x 10-3 75 Fair 
Quarry 2.73 x 10-2 48 Fair 
T
 ex
 as
  
2% FC 3.67 x 10-3 218 Poor 
4% FC 2.71 x 10-3 263 Poor 
8% FC 4.22 x 10-4 1262 Poor 
14% FC 7 x 10-5 3768 Poor 
Quarry 6.57 x 10-3 77 Fair 
Keystone Quarry 3.92 x 10-3 138 Fair 
Bladensburg Quarry 1.28 x 10-2 35 Fair 
Churchville Quarry 1.48 x 10-3 430 Poor 
Chantilly Quarry 5.66 x 10-4 1121 Poor 
Havre de Grace Quarry 1.48 x 10-2 33 Fair 
SHA lower limit 3.19 x 10-2 140 Fair 
SHA upper limit 3.84 x 10-3 29 Good 
 
 
  
 43 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FIGURES 
  
 44 
 
 
 
 
 
0
 20
 40
 60
 80
 100
 0.010.1110100
 Rockville
 Churchville
 Texas
 Bladensburg
 Chantilly 
Havre de Grace
 Keystone
 SHA Upper Limit
 SHA Lower LimitP
 er
 ce
 nt
  fi
 ne
 r
 #2
 00
 #3
 0
 3/
 4
 1 
1/
 2
 1/
 2
 3/
 8 #4
 Particle size (mm)
 (a)
  
0
 20
 40
 60
 8
 10
 0.010.1110100
 RCA--Plant A
 RCA--Plant B
 SHA Upper Limit
 SHA Lower Limit
 P
 er
 ce
 nt
  fi
 ne
 r
 Paticle size (mm)
 #2
 00
 #3
 0
 3/
 4
 1 1
 /2
 1/
 2
 3/
 8 #4
 (b)
  
 
FIGURE 1: Gradation of (a) GAB, and (b) RCA materials. 
 
 45 
 
 
0
 20
 40
 60
 80
 100
 0.010.1110100
 SHA lower limit
 SHA  upper limit
 Field
 retrieved
 Original quarry
  P
 er
 ce
 n
 t fin
 er
 Particle size (mm)
 1/
 12 3/
 8
 #4 #3
 0
 #2
 00
 3/
 4
 Rockville
  
0
 20
 40
 60
 80
 100
 0.010.1110100
 SHA lower limit
 SHA  upper limit
 Field
 retrieved
 Original quarry
  P
 er
 ce
 n
 t 
fi
 n
 er
 Particle size (mm)
 1/
 12 3/
 8
 #4 #3
 0
 #2
 00
 3/
 4
 Bladensburg
  
0
 20
 40
 60
 80
 100
 0.010.1110100
 SHA lower limit
 SHA  upper limit
 Field
 retrieved
 Original quarry
  P
 er
 ce
 n
 t fin
 er
 Particle size (mm)
 1/
 12 3/
 8
 #4 #3
 0
 #2
 00
 3/
 4
 Havre de Grace
  
0
 20
 40
 60
 80
 100
 0.010.1110100
 SHA lower limit
 SHA  upper limit
 Field
 retrieved
 Original quarry
  P
 er
 ce
 n
 t fin
 er
 Particle size (mm)
 1/
 12 3/
 8
 #4 #3
 0
 #2
 00
 3/
 4
 Texas
  
0
 20
 40
 60
 80
 100
 0.010.1110100
 SHA lower limit
 SHA  upper limit
 Field
 retrieved
 Original quarry
  P
 er
 ce
 n
 t fin
 er
 Particle size (mm)
 1/
 12 3/
 8
 #4 #3
 0
 #2
 00
 3/
 4
 Chantilly
  
 
 
FIGURE 2: Gradation of field retrieved samples. 
 
 46 
 
 
 
0
 100
 200
 300
 400
 500
 600
 700
 0 100 200 300 400 500 600 700 800
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 Keystone
 Rockville
 Texas
 R
 esi
 lie
 n
 t 
m
 o
 d
 u
 lu
 s 
(M
 P
 a)
 Bulk stress (kPa)
  
 
FIGURE 3: GAB resilient moduli at different loading sequences  
 47 
 
 
     
50
 100
 150
 200
 250
 300
 0 2 4 6 8 10
 Rockville
 Texas
 Bladensburg
 Keystone
 Havre de Grace
 Chantilly
 Churchville
 S
 M
 R
  (MPa
 )
 Moisture Content (%)
 FIGURE 4: Effect of moisture content on SMR values of GABs. 
  
 
 48 
 
 
0
 50
 100
 150
 200
 0 3 6 9 12 15
 SM
 R
  (M
 Pa
 )
 Fines content (%)
 (a)
  
 
 
40
 80
 120
 160
 200
 1.4 1.5 1.6 1.7 1.8 1.9
 SM
 R (
 M
 Pa
 )
 Gravel/Sand Ratio
 (b)
  
 
 
FIGURE 5:  Effect of (a) fines content, and (b) gravel-to-sand ratio  
on SMR of Rockville GAB. 
  
 49 
 
 
 
 
 
 
 
 
 
Large G/S       Optimum G/S   Small G/S 
 
(a)                                                 (b)                                                 (c) 
 
 
 
FIGURE 6: Arrangement of particles in a soil matrix with the variation of fines. 
(a) No or small fines content (large G/S ratio),  (b) dense graded (optimum G/S), and (c) 
high fines content (small G/S) (After Yoder and Witczak 1975, and Xiao et al. 2012) 
  
 50 
 
 
0
 100
 200
 300
 400
 500
 600
 700
 800
 0 100 200 300 400 500 600 700
 75B25BL
 50B50BL
 25B75BL
 RCA--Plant B
  Bladensburg (B)
 Re
 si
 lien
 t m
 od
 ul
 us (
 M
 Pa
 )
 Bulk stress (kPa)
 0
 100
 200
 300
 400
 500
 600
 700
 0 100 200 300 400 500 600 700
 RCA--Plant B
 75B25R
 50B50R
 25B75R
  Rockville (R)
 Resilient m
 od
 ul
 us (
 M
 Pa
 )
 Bulk stress (kPa)
 0
 100
 200
 300
 4 0
 5 0
 600
 700
 0 100 200 300 400 500 600 700
 75A25BL
 50A50BL
 25A75BL
 RCA--Plant A
 Bladensburg (B)
 Re
 si
 lie
 nt
  m
 od
 ulu
 s 
(M
 Pa
 )
 Bulk stress ( kPa)
 0
 100
 200
 300
 400
 500
 600
 700
 0 100 200 300 400 500 600 700
 RCA-Plant A
 75A25R
 50A50R
 25A75R
 Rockville (R)
 Re
 si
 lie
 nt
  m
 od
 ulu
 s 
(M
 Pa
 )
 Bulk stress (kPa)
  
FIGURE 7: Resilient moduli of recycled aggregate A and B, and their mixtures with two GABs.  
 51 
 
 
 
 
0
 0.02
 0.04
 0.06
 0.08
 0.1
 0 2000 4000 6000 8000 10000
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 Keystone
 Rockville
 Texas
 Plasti
 c 
st
 ra
 in
  (
 %
 )
 Number of repeated load cycles
  
 
 
 
FIGURE 8: Plastic strain of GABs under repeated load cycles.  All specimens are 
compacted at OMC. 
  
 52 
 
 
 
0
 0.01
 0.02
 0.03
 0.04
 0.05
 0.06
 0.07
 0 2000 4000 6000 8000 10000
 OMC
 OMC-2%
 OMC+2%
 Pla
 s
 ti
 c
  st
 ra
 in
  (
 %
 )
 Number of repeated load cycles
  
0
 0.01
 0.02
 0.03
 0.04
 0.05
 0.06
 0.07
 200 400 60 80 10000
 OMC
 OMC-2%
 OMC+2%
 Pl
 a
 stic s
 tr
 ai
 n
  (%
 )
 Number of repeated load cycles
  0
 0.01
 0.02
 0.03
 0.04
 0.05
 0.06
 0.07
 0 2000 4000 6000 8000 10000
 OMC
 OMC-2%
 OMC+2%
 P
 la
 sti
 c
  s
 tr
 a
 in
  (%
 )
 Number of repeated load cycles
  
0
 0.01
 0.02
 0.03
 0.04
 0.05
 0.06
 0.07
 0 2000 4000 6000 8000 10000
 OMC
 OMC-2%
 OMC+2%
 P
 la
 sti
 c
  s
 tr
 a
 in
  (%
 )
 Number of repeated load cycles
  
 
FIGURE 9: Effect of moisture content on plastic strain of the GABs   
(OMC=optimum compaction moisture content). 
Rockville Texas 
Churchville Chantilly 
 53 
 
 
 0
 0.2
 0.4
 0.6
 0.8
 1
 0 2000 4000 6000 8000 10000
 75A25R
 50A50R
 25A75R
 RCA-Plant A
 Rockville (R)
 Pla
 st
 ic
  st
 ra
 in
  (
 %
 )
 Number of repeated load cycles
 RCA-Plant A
  0
 0.1
 0.2
 0.3
 0.4
 0 200 400 600 800 1000
 75B25R
 50B50R
 25B75R
 RCA-Plant B
 Rockville (R)
 Pl
 astic st
 rai
 n
  (%
 )
 Number of repeated load cycles
 RCA-Plant B
  
 
FIGURE 10: Plastic strain of recycled concrete aggregates A and B, and their mixtures 
with Rockville GAB  
 
 54 
 
 
0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 0.010.1110100
 Original quarry
 2% fines
 6% fines
 8% fines
 10% fines
 12% fines
 14% fines
 SHA Upper Limit 
SHA Lower Limit
 P
 e
 rc
 ent
  fi
 n
 e
 r
 Particle size (mm)
 #
 2
 0
 0
 #
 3
 0
 3
 /4
 1 1
 /2
 1/
 2
 3
 /8 #
 4
 Rockville
 (a)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 0.010.1110100
 Original quarry
 2% fines
 4% fines
 6% fines
 8% fines
 10% fines
 12% fines
 14% fines
 16% fines
 SHA Upper Limit 
SHA Lower Limit
 P
 e
 rc
 ent
  fin
 e
 r
 Particle size (mm)
 #2
 0
 0
 #
 3
 0
 3/
 4
 1
  1
 /2
 1/
 2
 3/
 8
 #
 4
 Rockville
 (b)
  
0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 0.010.1110100
 Original quarry
 2% fines
 4% fines
 6% fines
 8% fines
 10% fines
 12% fines
 14% fines
 16% fines
 SHA Upper Limit 
SHA Lower Limit
 Per
 ce
 n
 t 
fi
 n
 er
  
Particle size (mm)
 #2
 0
 0
 #3
 0
 3/
 4
 1 1
 /2
 1/
 2
 3/
 8 #4
 Texas
 (c)
  
 
FIGURE 11: Change in gradations of (a) Rockville GAB due to adjustment between 0.6 
mm and 4.75 mm sieves, and (b) Rockville and (c) Texas GABs due to 
adjustment between 0.375 mm and 0.75 mm sieves.  
 55 
 
 
1x10
 -3
 1x10
 -2
 1x10
 -1
 0 2 4 6 8 10 12 14 16
 H
 yd
 ra
 u
 lic
  co
 n
 d
 u
 c
 ti
 v
 ity
  (c
 m
 /s
 )
  Fines Content  (%)
 (a)
 Rockville
  
1x10
 -3
 1x10
 -2
 1x10
 -1
 0 2 4 6 8 10 12 14 16
 H
 ydr
 a
 u
 li
 c
  c
 ond
 u
 cti
 v
 it
 y
  (c
 m
 /s
 )
 Fines content (%)
 SHA upper limit for gradation
 SHA lower limit for gradation
 Rockville
 (b)
  1x10
 -5
 1x10
 -4
 1x10
 -3
 1x10
 -2
 1x10
 -1
 0 2 4 6 8 10 12 14 16
 H
 yd
 ra
 u
 lic c
 o
 n
 d
 u
 c
 ti
 v
 it
 y (c
 m
 /s
 )
 Fines content (%)
 SHA upper limit
  for gradation
 SHA lower limit for gradation
 Texas
 (c)
  
FIGURE 12: Effect of fines on hydraulic conductivity of (a) Rockville GAB due to 
adjustment between 0.6 mm and 4.75 mm sieves, and (b) Rockville and (c) Texas GABs 
due to adjustment between 0.375 mm and 0.75 mm sieves.  
 56 
 
 
5x10
 -3
 1x10
 -2
 1.5x10
 -2
 2x10
 -2
 2.5x10
 -2
 3x10
 -2
 3.5x10
 -2
 45 50 55 60 65 70
 Hydr
 au
 lic
  co
 nduct
 iv
 ity (
 cm/
 s)
  Gravel content (%)
 (a)
 5 x 10
 -5
 1x10
 -3
 2x10
 -3
 3x10
 -3
 4x10
 -3
 35 40 45 50 55 60
 H
 ydr
 au
 lic
  c
 ond
 uc
 tiv
 ity (c
 m
 /s
 )
  Gravel content (%)
 (b)
  
FIGURE13: Effect of gravel content on hydraulic conductivity of  
(a) Rockville, and (b) Texas GABs 
  
 57 
 
 
 
5x10
 -3
 1x10
 -2
 1.5x10
 -2
 2x10
 -2
 2.5x10
 -2
 3x10
 -2
 3.5x10
 -2
 1.4 1.5 1.6 1.7 1.8 1.9
 H
 yd
 ra
 u
 lic
  c
 on
 du
 ctivity (
 cm/
 s)
 Gravel/sand ratio
 (a)
 5 x 10
 -5
 1x10
 -3
 2x10
 -3
 3x10
 -3
 4x10
 -3
 0.8 0.85 0.9 0.95 1 1.05 1.1 1.15
 Hy
 d
 ra
 u
 lic
  co
 n
 d
 u
 ctivit
 y 
(c
 m/
 s)
 Gravel/sand ratio
 (b)
  
FIGURE14: Effect of gravel/sand ratio on hydraulic conductivity of  
(a) Rockville, and (b) Texas GABs 
 58 
 
 
0
 1x10
 -2
 2x10
 -2
 3x10
 -2
 4x10
 -2
 0 2 4 6 8 10 12
 D
 10
 D
 30
 D
 60
 D
 50
 Hy
 d
 ra
 u
 li
 c co
 n
 duct
 ivity (
 c
 m/s
 )
 Particle diameter, D (mm)
 R
 2
 = 0.89
 R
 2
 = 0.95
 R
 2
 = 0.76
 R
 2
 = 0.84
 Rockville 
(a)
  
0
 1x10
 -3
 2x10
 -3
 3x10
 -3
 4x10
 -3
 5x10
 -3
 0 2 4 6 8
 D
 30
 D
 10
 D
 50
 D
 60
 Hy
 dr
 au
 lic
  c
 on
 ductivit
 y 
(c
 m
 /s
 )
 Particle diameter, D (mm)
 R
 2
 = 0.92
 R
 2
 = 0.89
 R
 2
 = 0.87
 R
 2
 = 0.77
 Texas
 (b)
  
 
FIGURE 15:  Effect of characteristic grain sizes on hydraulic conductivity of (a) 
Rockville, and (b) Texas GABs. 
 
 59 
 
 
0
 50
 100
 150
 200
 250
 0 50 100 150 200 250 300 350
 y = 0.42795x 
LW
 D m
 od
 ul
 us (
 M
 P
 a)
 Geogauge Young's modulus (MPa)
 r = 0.81
 t = 9.6 > t
 cr
  = 2.011
 0
 50
 100
 150
 200
 250
 0 5 0 15 20 25 30 35 40
 y = 3.7376x 
LW
 D m
 od
 ul
 us (
 M
 P
 a)
 Geogauge stiffness (MN/m)
 r = 0.81
 t = 9.6 > t
 cr
  = 2.011
  10
 15
 20
 25
 30
 140 150 160 170 180 190 200 210 220
 y = 0.11506x 
M
 ean
  ge
 oga
 ug
 e 
st
 iff
 ness 
(MN/m
 )
  
Mean SM
 R
 (MPa)
 r = 0.95
 t = 5.3 > t
 cr
  = 3.182
 100
 120
 140
 160
 180
 200
 220
 240
 100 120 4 160 180 200 220
 y = 0.98693x 
M
 ean geo
 gaug
 e 
Yo
 un
 g'
 s 
m
 od
 ul
 us (
 M
 Pa
 )
 Mean SM
 R 
(MPa)
 r = 0.91
 t = 3.8 > t
 cr
  = 3.182
  
40
 50
 60
 70
 80
 90
 100
 110
 120
 40 60 80 100 120 140
 y = 0.77992x 
M
 ean L
 W
 D m
 od
 ulu
 s 
(M
 Pa
 )
  Mean M
 R
  @ LWD Bulk Stress (MPa)
 Texas
 Rockville
 Havre de Grace
 Chantilly
 Bladensburg
 r = 0.97
 t = 6.9 > t
 cr
  = 3.182
  
 
FIGURE 16:  Comparison of laboratory and field moduli. 
 
 
 60 
 
 
1 x 10
 -4
 1 x 10
 -3
 1 x 10
 -2
 1 x 10
 -1
 1 x 10
 0
 Havre 
de Grace
 Rockville Chantilly Bladensburg Texas
 Laboratory test--Quarry samples
 Laboratory test--Field retrieved samples
 Field test
 H
 yd
 ra
 u
 lic co
 n
 d
 u
 ct
 ivi
 ty
  (
 cm
 /s
 )
  
 
FIGURE 17:  Comparison of laboratory and field hydraulic conductivities of GAB 
materials. 
  
 61 
 
 
 
 
10
 20
 30
 40
 50
 60
 70
 0.50.60.70.80.911.11.21.3
 a
 2
  = 0.08
 a
 2 
= 0.10
 a
 2 
= 0.12
 a
 2
  = 0.14
 R
 eq
 u
 ir
 ed
  ba
 se
  t
 h
 ic
 kn
 es
 s 
(c
 m
 )
 Base course drainage modification factor (m
 2
 )
  
 
FIGURE 18: Variation of required base thickness with the change in layer coefficient and 
drainage modification factor. 
  
 62 
 
 
    
25
 75
 125
 175
 225
 275
 1x10
 -3
 1x10
 -2
 1x10
 -1
 1x10
 0
 R
 eq
 u
 ir
 ed
  b
 as
 e 
th
 ick
 n
 es
 s (cm
 )
 Hydraulic conductivity (cm/s)
  
                                                               
 
FIGURE 19: Variation in required base thickness with base course hydraulic conductivity  
 
(a) 
 63 
 
 
0
 200
 400
 600
 800
 1000
 1200
 0 20 40 60 80 100
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Barber and Sawyer Method
 (a)
 (H=0.3 m)
 Rockville
  
0
 200
 400
 600
 800
 1000
 1200
 0 20 40 60 80 100
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Casagrande and Shannon Method
 (b)
 (H=0.3 m)
 Rockville
  
 
FIGURE 20: Variation in time to drain with percent drainage and fines content for 
Rockville GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon 
Method. H= 0.3 m was used throughout the analysis 
 64 
 
 
0
 50
 100
 150
 200
 0 10 20 30 40 50 60 70
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Barber and Sawyer 
Method
 (a)
 Rockville
  
0
 50
 100
 150
 200
 250
 0 10 20 30 40 50 60 70
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 im
 e 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Casagrande and Shannon
  Method
 (b)
 Rockville
  
FIGURE 21: Variation in time to drain with base thickness and fines content for 
Rockville GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon 
Method.  
 65 
 
 
0
 1000
 2000
 3000
 4000
 5000
 6000
 7000
 0 20 40 60 80 100
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Barber and Sawyer Method
 (a)
 (H=0.3 m)
 Texas
  
0
 1000
 2000
 3000
 4000
 5000
 6000
 7000
 0 20 40 60 80 100
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Casagrande and Shannon Method
 (b)
 (H=0.3 m)
 Texas
  
FIGURE 22: Variation in time to drain with percent drainage and fines content for Texas 
GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon Method. H= 0.3 
m was used throughout the analysis 
 66 
 
 
0
 1000
 2000
 3000
 4000
 5000
 6000
 7000
 0 10 20 30 40 50 60 70
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  a
 t 
U
 =50%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Barber and Sawyer 
Method
 (a)
 Texas
  
0
 1000
 2000
 3000
 4000
 50 0
 6000
 7000
 0 10 20 30 40 50 60 70
 2% fines
 4% fines
 8% fines
 14% fines
 SHA upper limit
 SHA lower limit
 T
 im
 e 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Casagrande and Shannon
  Method
  (b)
 Texas
  
 
FIGURE 23: Variation in time to drain with base thickness and fines content for Texas 
GAB: (a) Barber and Sawyer Method, and (b) Casagrande and Shannon Method.  
  
 67 
 
 
 
20
 30
 40
 50
 60
 70
 80
 40
 60
 80
 100
 120
 140
 5x10
 -3
 1x10
 -2
 1.5x10
 -2
 2x10
 -2
 2.5x10
 -2
 3x10
 -2
 3.5x10
 -2
 Time to drain    
Base thickness
 T
 im
 e 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 R
 eq
 u
 ired
  b
 ase th
 ickn
 ess 
(cm
 )
 Hydraulic conductivity (cm/s)
 14% fines
 8% fines
 4% fines
 2% fines
  
 
 
0
 500
 1000
 1500
 2000
 2500
 3000
 3500
 4000
 200
 400
 600
 800
 1000
 1200
 1400
 1600
 1800
 5 x 10
 -5
 1x10
 -3
 2x10
 -3
 3x10
 -3
 4x10
 -3
 5x10
 -3
 Time to drain
 Req.base Thickness
 T
 im
 e 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 R
 eq
 u
 ired
  b
 ase th
 ickn
 ess 
(cm
 )
 Hydraulic conductivity (cm/s)
 Texas 
2% Fines 
8% Fines 
14% Fines 
4% Fines 
(b)
  
 
FIGURE 24: Influence of hydraulic conductivity on required base thickness (Moulton 
method) and time-to-drain of highway base layers constructed with (a) Rockville, and (b) 
Texas GAB. 
 68 
 
 
0
 200
 400
 600
 800
 1000
 1200
 0 20 40 60 80 100
 Rockville
 Texas
 Keystone
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Barber and Sawyer 
Method
 (a)
 (H=0.3 m)
  
0
 200
 400
 600
 800
 1000
 1200
 0 20 40 60 80 100
 Rock ille
 Texas
 Keystone
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  (
 h
 o
 u
 rs
 )
 Drainage (%)
 Casagrande and
  Shannon Method
 (b)
 (H=0.3 m)
  
FIGURE 25: Variation in time to drain with percent drainage and GAB type: (a) Barber 
and Sawyer Method, and (b) Casagrande and Shannon Method. H= 0.3 m was used 
throughout the analysis 
 
 69 
 
 
0
 500
 1000
 1500
 0 10 20 30 40 50 60 70
 Rockville 
Texas
 Keystone
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Barber and Sawyer 
Method
 (a)
  
0
 500
 1000
 1500
 2000
 0 10 20 30 40 50 60 70
 Rockvi le 
Texas
 Keystone
 Bladensburg
 Churchville
 Chantilly
 Havre de Grace
 SHA upper limit
 SHA lower limit
 T
 ime 
to
  d
 ra
 in
  a
 t 
U
 =5
 0%
  (
 h
 o
 u
 rs
 )
 Base thickness (cm)
 Casagrande 
and Shannon 
Method
 (b)
  
 
 
FIGURE 26: Variation in time to drain with base thickness and  GAB type: (a) Barber 
and Sawyer Method, and (b) Casagrande and Shannon Method.  
 
 70 
 
 
0
 100
 200
 300
 400
 500
 600
 700
 1x10
 -4
 1x10
 -3
 1x10
 -2
 1x10
 -1
 R
 eq
 u
 ir
 ed
  b
 as
 e 
th
 ic
 kn
 es
 s 
(cm
 )
 Hydraulic conductivity (cm/s)
 Rockville
 Havre de Grace 
Texas
 Keystone
 Churchville
 Chantilly
 Bladensburg
  
0
 200
 400
 600
 800
 1000
 1200
 1x10
 -4
 1x10
 -3
 1x10
 -2
 1x10
 -1
 T
 ime 
to
  d
 ra
 in
  a
 t 
U
  =
  5
 0%
  (
 h
 o
 u
 rs
 )
 Hydraulic conductivity (cm/s)
 Rockville
 Havre de 
Grace 
Texas
 Keystone
 Churchville
 Chantilly
 Bladensburg
 Thickness of
  base (H) = 0.3m 
 
 
 
 
FIGURE 27: Effect of GAB hydraulic conductivity on (a) required base thickness, and 
(b) time to drain at 50% drainage.  
 
 
 71 
 
 
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 4
 10 7 1 0.08
 Poor Fair Good Excellent
 a
 2
  =0.08
 a
 2
 =0.10
 a
 2
  = 0.12
 a
 2
 =0.14
 Time to drain water from the base layer (days)
 Quality of Drainage
 C
 o
 s
 t 
o
 f 
G
 A
 B
  l
 ay
 e
 r 
 X
  1
 0
 5
  $
  
FIGURE 28: Effect of time to drain and quality of drainage on the cost of GAB layer. 
 
 
 
 
 
  
 72 
 
 
 
 
 
 
 
 
 
APPENDIX 
  
 73 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX A 
 
 
 
 
COMPACTION AND RESILIENT MODULUS TESTING  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 74 
 
 
 
 
FIGURE A.1: Particle degradation of GAB materials due to vibratory and impact compaction: (a) Texas (b) Rockville (c) 
Bladensburg (d) Churchville 
0
 20
 40
 60
 80
 100
 0.010.1110100
 P
 er
 ce
 nt
  F
 in
 er
 Particle Size, mm
 Virgin
 Vibratory Comp.
 Impact Comp.
 (a)
 0
 20
 40
 60
 80
 100
 0.010.1110100
 P
 er
 ce
 nt
  F
 in
 er
 Particle Size, mm
 Virgin
 Vibratory Comp.
 Impact Comp.
 (b)
 0
 20
 40
 60
 80
 100
 0.010.1110100
 P
 er
 ce
 nt
  F
 in
 er
 Particle Size, mm
 Virgin
 Vibratory Comp.
 Impact Comp.
 (c)
 0
 20
 40
 60
 80
 100
 0.010.1110100
 P
 er
 ce
 nt
  F
 in
 er
 Particle Size, mm
 Virgin
 Vibratory Comp.
 Impact Comp.
 (d)
 75 
 
 
 
 
 
Table A.1 AASHTO T 307-99 resilient modulus testing sequence for base and 
subbase materials. 
Sequence 
No. 
Confining 
Pressure, 
S3 (kPa) 
Maximum 
Axial 
Stress, Smax 
(kPa) 
Cyclic 
Stress, 
Scyclic 
(kPa) 
Constant 
Stress, 
0.1 Smax 
(kPa) 
Cycles 
0 103.4 103.4 93.1 10.3 500 
1 20.7 20.7 18.6 2.1 100 
2 20.7 41.4 37.3 4.1 100 
3 20.7 62.1 55.9 6.2 100 
4 34.5 34.5 31 3.5 100 
5 34.5 68.9 62 6.9 100 
6 34.5 103.4 93.1 10.3 100 
7 68.9 68.9 62 6.9 100 
8 68.9 137.9 124.1 13.8 100 
9 68.9 206.8 186.1 20.7 100 
10 103.4 68.9 62 6.9 100 
11 103.4 103.4 93.1 10.3 100 
12 103.4 206.8 186.1 20.7 100 
13 137.9 103.4 93.1 10.3 100 
14 137.9 137.9 124.1 13.8 100 
15 137.9 275.8 248.2 27.6 100 
 
 
 
 
 
 
 
 
 
 
 76 
 
 
STEP-BY-STEP RESILIENT MODULUS TEST PROCEDURE 
1) Turn on Geocomp Load Trac II  
2) Turn the air pressure pump on 
3) Measure the specimen height and diameter 
4) Place the porous stone on bottom plate 
5) Place the filter paper on bottom porous stone 
6) Place the specimen on bottom plate 
7) Place the filter paper on top of the specimen 
8) Place the porous stone on filter paper 
9) Place the top plate on top of the specimen 
10) Place rubber membrane over specimen using a mold 
11) Place two O- rings on both bottom and top of the plates to hold the membrane in 
place 
12) Plug the drainage tubes on top plate. 
13) Place the cell on bottom cap 
14) Place cover plate, it should not be tight 
15) Place LVDT on top of chamber 
16) Screw cover plate with three rods carefully 
17) Plug air supply hose into cell 
18) Log into PC and open the Resilient modulus RM version 5.0 software 
19) Input Specimen height, diameter, and weight. 
20) Input the loading and pressure data which is designed for base and subbase test 
protocol 
21) Click on the load calibration menu and check the applied load with the load data that 
you entered 
22) Click run test and save the file. 
  
 77 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX B 
 
 
 
 
PICTURES OF EQUIPMENT USED IN THE CURRENT RESEARCH STUDY 
 
  
 78 
 
 
 
 
 
 
                                                                         
FIGURE B.1 (a) Resilient Modulus Testing Equipment (b) Vibratory compactor. 
  
(b) 
(a) 
 79 
 
 
Filling Port
 Screens 
and 
Support 
Plates
 Bubble 
Tube
 Inlet 
Reservoir
 Coupling
 Soil 
Specimen
 Drain Port
 Water 
Bath
 H
  
 
 
FIGURE B-2: Bubble tube constant head permeameter 
  
 80 
 
 
 
 
 
 
FIGURE B.3: Nuclear density gauge for compaction and moisture content testing. 
  
 81 
 
 
 
 
 
 
FIGURE B.4: Light weight deflectometer 
 
 82 
 
 
    
 
 
FIGURE B.5: Geogauge used in field testing (After Humboldt Mfg.Co) 
 
 
 
 
 
 
 83 
 
 
 
 
FIGURE B.6:  Field hydraulic conductivity test on GAB. (ASTM D6391--Stage 1) 
  
 84 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX C 
 
 
 
 
FIELD TEST RESULTS 
85 
 
 
Table C.1  Field Tests Results. 
 
Quarry 
Name 
Geogauge LWD 
Nuclear Density 
Gauge 
Young's 
Modulus 
(MN/m) 
Stffness 
(MN/m) 
Modulus(MN/
 m?) 
d  
(kN/m3) 
w (%) 
RockV1 223.45 22.16 70.42 22.59 4.7 
RockV2 224.80 25.92 84.99 21.00 6.4 
RockV3 232.67 26.82 81.52 21.43 7.1 
RockV4 189.26 21.81 67.87 22.21 4.7 
RockV5 183.15 21.11 50.17 21.99 4.3 
RockV6 212.34 24.47 70.09 21.46 4 
RockVII1 216.81 24.99 89.02 22.98 0.2 
RockVII2 212.42 24.49 110.29 23.48 0.7 
RockVII3 232.28 26.77 87.98 22.38 1 
Texas1   35.22 174.11 20.73 4.4 
Texas2 204.07 25.21 132.49 20.66 4.8 
Texas3   33.59 127.27 22.06 3.6 
Texas4 317.57 36.61 216.05 22.79 2.9 
Texas5 283.87 32.72 168.65 22.89 2.9 
HavdG1 76.20 8.78 57.03 23.56 6.2 
HavdG2 95.73 26.00 26.15 23.17 5.8 
HavdG3 139.16 16.04 40.48 22.93 6.7 
HavdG4 63.26 7.29 20.85 23.50 7.1 
HavdG5 194.73 22.45 85.81 23.37 6.3 
HavdG6 146.96 16.94 22.59 23.26 6.7 
HavdG7 43.34 5.00 15.99 23.00 6.8 
HavdGSo1 153.64 17.71 84.89 23.03 6.2 
HavdGSo2 147.65 17.02 41.42 23.61 5.5 
HavdGSo3 149.53 17.24 59.95 22.89 5.7 
HavdGSo4 89.01 10.26 31.35 23.14 5.2 
HavdGSo5 190.46 18.32 67.45 22.93 6.2 
HavdGSo6 156.14 18.00 46.13 24.00 5.4 
HavdGSo7 113.96 13.14 17.36 23.00 6.8 
HavdGSo8 114.63 13.22 20.70 23.00 4.5 
HavdGSo9 188.36 21.60 115.58 23.34 7.8 
HavdGSo10 207.95 23.97 118.47 22.84 7.1 
HavdGSo11 122.60 14.13 86.52 22.76 6.5 
HavdGSo12 119.22 13.75 58.49 22.92 6.3 
Tex695-1 138.16 15.92 75.96 22.46 5.1 
Tex695-2 131.94 15.21 71.66 21.49 4.6 
Tex695-3A 144.21 16.62 40.50 22.09 4.1 
86 
 
 
 
 
 
 
  
Tex695-3B 173.82 20.04 55.31 21.52 5.2 
Tex695-4 143.85 16.58 52.06 21.79 4.9 
Blaburg 1 137.27 15.83 56.33 22.87 5.2 
Blaburg 2 159.92 18.44 60.39 22.82 6.6 
Blaburg 3 167.46 19.31 54.58 22.70 6.8 
Blaburg 4 203.05 23.41 89.43 22.29 6 
Blaburg 5 169.45 19.54 51.32 23.64 5.7 
Blaburg 6 178.71 20.60 68.69 22.97 5.5 
Chantily1 128.77 14.84 53.29     
Chantily2 101.81 11.74 51.1 
 
  
Chantily3 198.54 23.07 75.11 
 
  
Chantily4 109.43 12.62 38.29 
 
  
Chantily5 131.45 15.15 29.8 
 
  
Chantily6 136.85 15.78 47.56     
87 
 
 
REFERENCES 
 
AASHTO Guide (1993). Guide for design of pavement structures. American 
Association of State Highway and Transportation Officials, Washington D.C. 
AASHTO Specification (2010). Standard specification for transportation materials and 
methods of sampling and testing. American Association of State Highway and 
Transportation Officials, Washington D.C. 
ASTM. Specification (2010).  Annual Books of ASTM Standards, Road and Paving 
Materials. American Standards of Testing Materials. 
Abhijit.P., Patil,.J. (2011). Effects of bad drainage on roads. Civil and Environmental 
Research,  IISTE, Vol 1, No.1. India 
Barber, E.S and Sawyer, C.L. (1952). Highway subdrainage: Proceedings of Highway 
Research Board.  pp 643-666. 
Bozyurt, O., Keene,A., Tinjum,J.M., Edil,T.B. (2012). Freez-thaw effects on stiffness 
of unbound recycled road base. Transportation Research Board Annual 
Meeting. Washington D.C. 
Bennert,T., Maher, A. (2005).  The development of a performance specification for 
granular base and subbase material. Federal Highway Authority, FHWA-NJ-
 2005-003. New Jersey. 
Cetin, B., Aydilek, A.H., Guney. Y.  (2010). Stabilization of recycled base materials 
with high carbon fly ash.  Resources, Conservation and Recycling 54. pp 878-
 892 
Casagrande, A., Shannon, W.L. (1952). Base course drainage for airport pavements. 
Proceedings of American Society of Civil Engineers.  Vol.77, pp.792-814. 
88 
 
 
Cedergren, H.R., Arman, J.A., O'Brien, K.H. (1973). Development of guidelines for 
the design of subsurface drainage systems for highway pavement structural 
sections. Federal Highway Administration. FHWA-RD-73-14. 
Christopher, B. R., and McGuffey, V.C. (1997). NCHRP synthesis of highway practice 
no.239: Pavement subsurface drainage systems. National Research Council, 
Washington, D.C.  
Dawson, A. R., Thom, N. H., and Paute, J. L.  (1996). Mechanical characteristics of 
unbound granular materials as a function of condition. Flexible Pavements, 
Proc., Eur. Symp.  Rotterdam, Netherlands, pp 35-44. 
Dawson, A.R. (2009). Introduction. in: Water in Road Structure. Ch. 1. Springer.  p 
436 
Diefenderfer.P., Abhijit.P., Patil,.J. (2001). Six adverse effects of bad drainage on 
roads. Civil and Environmental Research, Vol 1, No.1. India 
Erlingsson, S., Baltzer, S., Baena, J. and Bjarnason, G  (2009). Measurement 
techniques for water: water in road structures. Ch. 3. Springer.  p 436. 
Haynes, J. G., and Yoder, E. J.  (1963). Effects of repeated loading on gravel and 
crushed stone base course materials used in the aasho road test. Highway 
Research Record  39. 
Hicks, R. G. (1970). Factors influencing the resilient properties of granular materials. 
PhD Thesis. Berkeley: University of California.  
Hicks, R. G. and Monismith, C. L. (1971). Factors influencing the resilient properties 
of granular materials. Highway Research Record 345. pp 15-31. 
89 
 
 
Jorenby, B. N., and Hicks, R. G. (1986). Base course contamination limits. 
Transportation Research  Record 1095. pp 86-101. 
Kamal, M. A., Dawson, A. R., Farouki, O. T., Hughes, D. A. B. (1993). Field and 
laboratory evaluation of the mechanical behaviour of unbound granular 
materials in pavements. Transportation Research Record 1406. pp. 88-97. 
Kazmee, H., Tutumluer, E., Mishra,D. (2012). Effects of material blending on strength, 
modulus and deformation characteristics of recycled concrete aggregates. 
Transportation Research Board Meeting. Washington D.C 
Khogali, W. E., and Mohammed, E. H. (2004). Novel approach for characterization of 
unbound materials. Transportation  Research Record 1874. pp 38-44. 
Kolisoja, P. (1997). Resilient deformation characteristics of granular materials. PhD 
Thesis.  University of Technology, Tampere. 
 Kutay,M.E., Aydilek, A.H., Masad,E., Harman,T. (2007). Computational and 
experimental evaluation of hydraulic conductivity anisotropy in hot-mix 
asphalt. International Journal of Pavement Engineering, Vol. 8, No. 1, pp 29?
 43. 
Moulton, L.K. (1980). Highway subdrainage design. Federal Highway Administration, 
FHWA-TS-80-224. Washington D.C 
NCHRP 1-28A. (2003). Harmonized test methods for laboratory determination of 
resilient modulus for flexible pavement design. Asphalt concrete material. Vol. 
II 
NCHRP  626. (2009).  NDT technology for quality assurance of HMA pavement 
construction. Transportation Research Board. 
90 
 
 
NCHRP 10-65 (2006).  Nondestructive Testing technology for quality control and 
acceptance of flexible pavement Construction. Transportation Research Board. 
 
Poon, C.S., Shui, Z.H., Lam, L., Fok.H. (2004). Influence of moisture states of natural 
and recycled aggregates on the slump and compressive strength of concrete. 
Cement and Concrete Research. Vol.34, pp 31-36. 
Rada, G., and Witczak, M. W. (1981). Comprehensive evaluation of laboratory 
resilient moduli results for granular material. Transportation Research Record. 
810. pp 23-33. Washington D.C. 
Ridgeway, H.H. (1976). Infiltration of water through the pavement surface. 
Transportation Research Record 616,TRB, National Research Council, 
Washington D.C. pp 98-100. 
Robinson, R.G. (1974). Measurement of the elastic properties of granular materials 
using a resonance method.  TRRL Supplementary Report.No. 111UC,  
Smith, W. S., and Nair, K. (1973). Development of procedures for characterization of 
untreated granular base coarse and asphalt-treated base course materials. 
Federal Highway Administration,  FHWA-RD-74-61.Washington D.C. 
Sweere, G. T. H. (1990). Unbound granular basis for roads,  PhD Thesis. University of 
Delft, The Netherlands. 
Thom, N. H., and Brown, S. F. (1987).  Effect of moisture on the structural 
performance of a crushed-limestone road base.  Transportaton Research 
Record 1121. pp 50-56 Washington D.C.. 
91 
 
 
Siswosoebrotho,B.I., Widodo.P, Augusta.E. (2005) .The influence of fines content and 
plasticity on the strength and permeability of aggregate for base course 
material. Eastern Asia Society for Transportation Studies. Indonesia, Vol 5, pp 
845-856. 
Uzan, J. (1985). Characterization of granular material.  Transportation Research 
Record 1022, pp 52-59. Washington D.C. 
Uthus, L. (2007). Deformation properties of unbound granular aggregates,  PhD 
Thesis. Department of civil and transportation engineering,  Norwegian 
University of Science and Technology. 
Vuong, B. (1992). Influence of density and moisture content on dynamic stress-strain 
behavior of a low plasticity crushed rock. Road and Transportation Research, 
TR 1(2). pp 88-100.  
White, D., Bergeson, K., and Jahren, C. (2002). Embankment quality: phase III. 
Department of Civil and Construction Engineering, Iowa State University, final 
report, Iowa DOT Project TR-401, CTRE Project 97-08, 
Witczak, M.W. and Uzan, J. (1988).  The universal airport pavement design system. 
Report I of IV.  University of Maryland. 
 Xiao,Y., Tutumluer, E., Qian, Y. (2012). Gradation effects influencing mechanical 
properties of aggregate base and granular subbase in Minnesota. Transportaton 
Research Board Meeting. Washington D.C.